Physical AI Brief

Modern conversational agents condition on an ever-growing dialogue history at each turn, incurring redundant attention and encoding costs that grow with conversation length. Naive truncation or summarization degrades fidelity, while existing context compressors lack cross-turn memory sharing or revision, causing information loss and compounding errors in long dialogues. We revisit the context compression under conversational dynamics and empirically present its fragility. To improve both efficiency and robustness, we introduce Context-Driven Incremental Compression (C-DIC), which treats a conversation as interleaved contextual threads and stores revisable per-thread compression states in a single, compact dialogue memory. At each turn, a lightweight retrieve, revise, and write-back loop shares information across turns and updates stale memories, stabilizing long-horizon behavior. In addition, we adapt truncated backpropagation-through-time (TBPTT) to our multi-turn setting, learning cross-turn dependencies without full-history backpropagation. Extensive experiments on long-form dialogue benchmarks demonstrate superior performance and efficiency of C-DIC; notably, C-DIC shows stable inference latency and perplexity over hundreds of dialogue turns, supporting a scalable path to high-quality dialogue modeling.

General-purpose large language models (LLMs) are routinely used as baselines when evaluating specialized pathology models on whole-slide images (WSIs). Because WSIs exceed contemporary model context limits, LLM baselines routinely use small, high-magnification patches processed independently via majority voting, without systematic evaluation of seemingly inconsequential design choices such as patch size, patch count, and magnification. Generalist LLMs have consistently underperformed specialized systems, reinforcing the perception that domain-specific training or architectural adaptation is necessary for pathology tasks involving WSIs. Here, we conduct a systematic factorial analysis of four input design factors: inference mode, patch size, magnification, and patch count. We demonstrate that prior studies have overstated the gap between specialized models and general-purpose LLMs by choosing non-optimized input configurations. On the MultiPathQA benchmark, switching to a single balanced configuration (large patches at lower magnification, processed jointly) raises GPT-5 from 15.1% to 39.5% on cancer-type classification (TCGA) and from 38.1% to 62.9% on organ classification (GTEx). Per-task optimization yields further gains up to 43.9% (TCGA) and 71.6% (GTEx). The same configuration generalizes to two other models and to a fully held-out CPTAC cohort, where it improves Gemini 3 Flash by 23.4 percentage points without any task-specific tuning.

Contact-rich manipulation requires force sensitivity, but many robot arms lack dedicated force sensors due to their high cost. We present Neural External Torque Estimation (NEXT), a data-driven method that estimates external joint torques without needing any dedicated force sensors. NEXT trains in 1 minute from only 10 minutes of free-motion data, yet achieves estimates comparable to dedicated joint-torque sensors. NEXT enables force-feedback teleoperation on low-cost arms and improves policy learning through Force-Informed Re-Sampling Training (FIRST), which up-samples pre-contact and contact segments during behavior cloning. Across five long-horizon tasks, FIRST outperforms prior force-aware policies by over 17% in task progress. Together, NEXT and FIRST bring force-aware teleoperation and policy learning to off-the-shelf robots without additional sensing hardware. Video results and code are available at https://jasonjzliu.com/factr2

Vision-Language-Action (VLA) models inherit semantic grounding from large-scale pretraining and perform competently across in-distribution manipulation tasks. This grounding, however, is built on static image-text pairs, whereas manipulation is a continuous, contact-rich process whose dynamics this pretraining cannot capture. We present World Pilot, a VLA framework that augments the policy with priors from a World-Action Model (WAM), routed into the decision chain through two complementary pathways. Latent Steering conditions the perception layer on a scene-evolution latent, and Action Steering supplies an anticipated trajectory as a motion prior to the action generator. Together the two priors equip the VLA with an anticipated view of the scene and a trajectory-level motion hint alongside its semantic conditioning, and the scene-evolution prior remains effective even when supplied by a video-pretrained world model that has not been action-post-trained. World Pilot attains a state-of-the-art Total success rate of 84.7% on the LIBERO-Plus zero-shot OOD benchmark and the highest success rate on every real-robot setting across four manipulation tasks, with the largest margins under shifts in viewpoint, geometry, deformable state, and pose. Project Website: https://world-pilot.github.io/

Vision-Language Models (VLMs) are increasingly deployed as high-level planners for embodied agents, with an emerging strategy of scaling test-time compute to improve capability. However, we observe that doing so increases latency, token usage, and FLOPs while yielding uneven, often diminishing gains in downstream success, limiting where embodied agents can be deployed. We argue that choosing when and where to spend test-time compute is central to bringing frontier performance to the real world. We introduce DIRECT, a routing framework that uses multimodal scene context to allocate compute per prompt, improving the success--cost Pareto frontier over fixed model selection. Across three dominant scaling axes, namely chain-of-thought depth, model size, and memory history, our experiments on VLABench and RoboMME show that test-time compute is not a uniform lever: different axes yield qualitatively distinct capability gains. We validate these insights on a physical Franka arm in a DROID setup spanning zero-shot manipulation and long-horizon chaining, where our router matches or exceeds a stronger model's success rate at up to 65% lower average latency. Ultimately, our results show that naively scaling test-time compute is wasteful, and that DIRECT can provide frontier-level embodied planning in robotic systems at a fraction of the cost. Project page can be found at jadee-dao.github.io/direct/.

Long input sequences are central to document understanding and multi-step reasoning in Large Language Models, yet the quadratic cost of attention makes inference both memory-intensive and slow. Context distillation mitigates this by compressing contextual information into model parameters, and recent work such as Doc-to-LoRA amortizes context distillation into a single forward pass that generates one LoRA adapter per document. However, producing a single monolithic adapter for all queries leads to irrelevant-query interference, limited compositional recall, and poor scalability to long-document reasoning. To address these challenges, we propose Doc-to-Atom (Doc2Atom), a compositional parametric memory framework that decomposes each document into semantically typed knowledge atoms. Each atom is compiled into an independent micro-LoRA adapter and a provenance retrieval key. At inference time, a lightweight query router selects and assembles only the relevant atoms into a query-specific adapter, which is then injected into a frozen base model. The entire system is trained end-to-end through a multi-objective distillation framework. Experiments on six diverse QA benchmarks demonstrate that Doc2Atom outperforms Doc-to-LoRA baselines while reducing the memory cost of document internalization.

Vision-language-action (VLA) models can describe scenes and reason about them in language, yet still struggle to ground their actions in the dense 3D world around them. Existing approaches either inject features from a frozen 3D foundation model without an objective that ensures the policy uses them, or constrain geometry with sparse box and map losses that provide no dense spatial signal. We introduce VLGA, the first vision-language-action model supervised to reconstruct the dense 3D world it drives through. VLGA introduces geometry as a fourth modality alongside vision, language, and action through a dedicated expert supervised by a per-pixel pointmap regression loss against LiDAR. Extensive experiments conducted on challenging nuScenes and Bench2Drive datasets for open-loop and closed-loop evaluations, respectively, show the superiority of VLGA over counterpart VLA methods. In particular, on open-loop nuScenes, VLGA sets a new state of the art among VLA methods without ego status, with the lowest L2 (0.50\,m average) and 3-second collision rate (0.18\%). On closed-loop Bench2Drive, VLGA attains the state-of-the-art driving score of 79.08, +0.71 over the strongest prior VLA, at comparable efficiency and comfort.

Recently, large language models (LLMs) have achieved promising progress in the fields of classical Chinese translation and the generation of classical poetry. However, domain-specific research on precise translation and affective-semantic understanding of classical poetry remains limited. The main challenge is that most studies treat the poetic appreciation task as a general-domain problem, neglecting the distinctive features of poetic appreciation, while high-quality and domain-specific datasets are extremely limited. To address this limitation, we decompose the task into three subtasks: term interpretation, semantic interpretation, and emotional inference. Based on multiple open-source datasets, we perform data cleansing and alignment to construct the Classical Chinese Poetry Instruction Pair Dataset (CCPoetry-49K), which comprises 49,404 high-quality instruction-response pairs explicitly optimized for this domain. We then propose a domain-specialized LLM, called PoetryQwen, by applying Low-Rank Adaptation (LoRA) to fine-tune the Qwen2.5-14B model. Experimental results on the CCL25-Eval Task 5 benchmark demonstrate that PoetryQwen achieves a score of 0.757, representing a 9.7% improvement over the Qwen2.5-14B-Instruct baseline (0.690). These findings clearly indicate that PoetryQwen significantly enhances performance in precise translation and emotional understanding of classical poetry. We present new dataset and methodological considerations intended to support the domain-specific optimization of LLMs.

Large Language Models (LLMs) have democratized database access through Text-to-SQL, but moving from prototypes to production remains difficult. Real deployments must handle strict SQL dialects, massive schemas, and evolving user preferences, while supervised fine-tuning is costly and rigid and agentic test-time scaling is expensive. We present Tahoe, a system that treats prompt optimization as a dynamic data management problem. Tahoe uses an error-driven hint learning pipeline across Development and Deployment to consolidate debugging traces into a structured Hint Bank. Compiler feedback is distilled into reusable Syntax Hints for dialect-specific rules, while execution and user feedback are converted into Semantic Hints for schema- and user-specific logic. Tahoe further introduces a Strategy Layer that models conflicting user intents as competing strategies under shared natural-language triggers, with recency signals and post-learning attribution statistics that summarize empirical success, harm, inertness, and support. At inference time, Tahoe retrieves relevant hints and guides the LLM through Logic Planning followed by SQL Synthesis. We implement and evaluate the development-phase workflow, leaving deployment-time human-feedback updates for future work. On Spider 2.0-Snow, Tahoe substantially improves Text-to-SQL without updating model parameters. On 113 supervised Spider 2.0-Snow-0212 examples using GPT-5.5, Tahoe raises pass rate from 61.95 percent to 79.42 percent and pass-at-4 from 72.57 percent to 87.61 percent, achieves 100 percent Snowflake syntax pass rate, and reduces average compiler-feedback critic rounds from 2.79 to 0.12 per sampled candidate. The same Hint Bank also transfers to weaker backbones, including a 19.7 percentage-point pass-rate gain on Doubao-2.0-lite.

Advancing scientific understanding through mechanistic modeling requires posing the right experimental questions to yield maximally informative data. To automate this pursuit within cognitive science, we introduce ATLAS (Active Theory Learning for Automated Science), an active learning framework for the data-driven discovery of interpretable behavioral models. ATLAS iterates between generating mechanistic hypotheses--instantiated as a diverse ensemble of sparse neural networks (Disentangled RNNs)--and designing experiments that optimally distinguish between them. We test this approach on the problem of recovering reinforcement learning agents from their behavior in bandit tasks. ATLAS designs varied sequences of qualitatively novel experiments with temporal structure tailored to underlying agent characteristics. The models trained on these experiments are evaluated against a comprehensive set of metrics for mechanistic modeling that capture behavioral, structural, and computational similarity. ATLAS achieves a 5-10x improvement in sample efficiency across all metrics compared to random experimentation, and its performance is further validated against expert-designed experiments derived from literature. These in silico results showcase ATLAS's potential to accelerate human-interpretable insights in cognitive science and other domains where scientific inquiry relies on discovering mechanistic models.

Modern LLM training pipelines increasingly rely on other models to generate data, filter corpora, judge outputs, and guide development decisions. These dependencies are recursive: a model may depend on an upstream artifact whose own dependencies are documented only in separate releases and artifacts. As a result, the full dependency structure is fragmented across heterogeneous public artifacts, with complexity and recursive depth far outpacing humans' ability to trace. We introduce ModSleuth, an agentic system that recursively reconstructs LLM dependency graphs from public artifacts with source-grounded evidence. We find that the primary challenge is no longer information extraction, but defining what constitutes a dependency and reconciling artifact references across inconsistent documentation. We address these challenges through a formalization that distinguishes direct and indirect dependencies, represents heterogeneous pipeline roles through operation-centered relationships, and resolves artifact identities across names, versions, and repositories. Applying ModSleuth to four public-artifact-rich LLM releases, we recover 1,060 source-verified dependencies and construct large-scale dependency graphs of modern LLM development. These graphs reveal multi-hop license obligations, train-evaluation coupling, discrepancies between released and training-time artifacts, and documentation inconsistencies that would otherwise be difficult to uncover. We release ModSleuth and the resulting dependency graphs to support transparent analysis of the increasingly complex ecosystems underlying modern LLMs.

Recent advances in agentic Reinforcement Learning (RL) have substantially improved the multi-turn tool-use capabilities of large language model agents. However, most existing methods assign credit over coarse heuristic units, such as tool-call boundaries or fixed workflows, making it difficult to identify which intermediate decisions influence downstream outcomes. In this work, we study agentic RL from two perspectives: \textit{where to branch and how to assign credit after branching}. Our pilot analysis shows that influential decision points are broadly distributed throughout the generated sequence rather than concentrated at tool calls, while token entropy alone does not reliably reflect their impact on final outcomes. Motivated by these observations, we propose \textbf{Agentic Procedural Policy Optimization (APPO)}, which shifts branching and credit assignment from coarse interaction units to fine-grained decision points in the sequence. APPO selects branching locations using a Branching Score that combines token uncertainty with policy-induced likelihood gains of subsequent continuations, enabling more targeted exploration while filtering out spurious high-entropy positions. It further introduces procedure-level advantage scaling to better distribute credit across branched rollouts. Experiments on 13 benchmarks show that APPO consistently improves strong agentic RL baselines by nearly 4 points, while keeping efficient tool-calls and maintaining behavior interpretability.

The Strength Pareto Evolutionary Algorithm 2 (SPEA2) is a popular and prominent evolutionary algorithm for solving multi-objective optimisation problems. Despite its popularity, theoretical analyses of SPEA2 have only appeared recently. Moreover, these analyses focus exclusively on how SPEA2 handles non-dominated solutions and disregard the algorithmic components responsible for handling dominated solutions. We conduct a first runtime analysis of SPEA2 for which these components are analysed. We prove that, unlike other prominent algorithms, including NSGA-II, NSGA-III and SMS-EMOA under the same setting of constant population size and duplicate elimination, SPEA2 is unable to cover the Pareto front of the OneTrapZeroTrap benchmark efficiently. Our results indicate that using k-th nearest-neighbour distance in the fitness assignment provides an insufficient signal to maintain diversity among dominated individuals. To address this issue, we propose an improved variant, SPEA2$^+$, that considers all pairwise distances. The new algorithm achieves the same performance guarantees as the other prominent algorithms on OneTrapZeroTrap, while matching the performance of the original SPEA2 on simpler problems. Experimental results complement our theoretical findings.

Reinforcement Learning (RL) with verifiable environments has emerged as a powerful approach for enhancing the reasoning capabilities of Large Language Models (LLMs). While prior research demonstrates that scaling environment quantity improves RL performance, existing manual or individual construction methods suffer from linear scaling limits, thereby hindering scalable reasoning generalization. This paper introduces RACES (\textbf{R}ecursive \textbf{A}utomated \textbf{C}omposition for \textbf{E}nvironment \textbf{S}caling), a framework that conceptualizes verifiable environments as composable building blocks that can be recursively assembled. The key insight is that when the codomain (output type) of one environment matches the domain (input type) of another, they can be automatically fused into a new verifiable environment, enabling recursive composition. RACES is implemented with 300 individual environments and defines a set of composition operators (\textsc{SEQUENTIAL}, \textsc{PARALLEL}, \textsc{SORT}, and \textsc{SELECT}) that induce diverse reasoning patterns. Extensive experiments show that RL training on these composite environments consistently enhances reasoning generalization. Specifically, RACES improves DeepSeek-R1-Distill-Qwen-14B by an average of 3.1 points (from 48.2 to 51.3) and boosts Qwen3-14B performance from 58.8 to 61.1 on six benchmarks, which are unseen during the construction of training environments. Moreover, RACES achieves performance comparable to training on 300 individual environments using only 50 base environments, demonstrating significant efficiency in environment utilization.

Human-in-the-loop reinforcement learning (HiL-RL) has emerged as an effective paradigm for real-world robotic manipulation, enabling online policy improvement with human guidance. However, current HiL-RL frameworks remain intervention-intensive, relying on frequent human corrections to redirect the policy out of unproductive exploration, which incurs high labor cost and limits real-world scalability. To address this, we propose UniIntervene, an agentic intervention model that detects unproductive exploration and autonomously recovers the policy toward high-value states, taking over the bulk of interventions from human operators. Specifically, UniIntervene first performs future-conditioned action-value estimation, predicting the latent consequence of the current action and evaluating its induced value, which provides a more stable progress signal. Building on this, a temporal value-risk critic aggregates recent value dynamics and triggers intervention when the estimated value exhibits sustained stagnation or degradation. When intervention is required, UniIntervene retrieves a high-value recovery target from a memory of past intervention episodes and produces executable corrective actions through a goal-conditioned recovery policy. In this way, UniIntervene turns intervention from passive human correction into a value-aware recovery process for efficient real-world RL. Extensive experiments on diverse real-world manipulation tasks demonstrate that UniIntervene improves the average success rate by 8.6% while reducing human interventions by 57% relative to state-of-the-art HiL-RL baselines.

Reinforcement learning (RL) has become a key component in modern large language models, yet the rollout stage remains the key bottleneck in RL training pipelines. Although Multi-Token Prediction (MTP) offers a natural solution to accelerate rollouts through speculative decoding, many studies have observed that MTP acceptance rates degrade significantly during RL training, leading to limited speedup performance. To address this bottleneck, we present Bebop, a systematic study of MTP in LLM post-training, and offer practical recipes to integrate MTP into large-scale RL pipelines. First, we reveal that the MTP acceptance rate is fundamentally bounded by the fluctuation of model entropy, which demonstrates a clear negative linear relationship with the rise of entropy in the RL stage. Second, we show that probabilistic rejection sampling largely alleviates the disturbance introduced by entropy in RL compared to greedy draft sampling. We further identify that the conventional MTP training objectives (cross-entropy or KL) are suboptimal in such settings, and therefore we propose a novel end-to-end TV loss that directly optimizes multi-step rejection sampling acceptance rate, yielding ~10% acceptance rate improvements, achieving up to 95% acceptance rates and up to 25% extra inference throughput gains across mathematical reasoning, code generation, and agentic tasks. Third, we test various online MTP training strategies during RL and show that pre-RL MTP training with e2e TV loss and rejection sampling achieves a consistent acceptance rate and speedup throughout the entire RL, eliminating the need for costly online MTP updating. We provide extensive experiments and analysis that validate our findings. Experimental results show our method achieves up to 1.8x end-to-end acceleration in async RL training of Qwen3.5, Qwen3.6, and Qwen3.7 models.

Vision-Language-Action (VLA) models that couple pretrained Vision-Language Models (VLMs) with continuous action experts have achieved strong manipulation performance, yet generalization to out-of-distribution (OOD) language instructions remains poor. A known challenge is the structural imbalance in VLA data, where language is far less diverse than visual and action content, making policies prone to visual shortcuts. While discrete-action methods mitigate this through vision-language co-training, continuous action experts lack such protection: they start from random initialization and learn entirely from imbalanced data, producing noisy gradients that corrupt the VLM and fail to exploit its language capability. We address this from a Bayesian perspective, factorizing the policy into a language-agnostic Vision-Action (VA) prior and a language-conditioned VLA likelihood, and propose APT, a two-stage training method emphasizing Action expert PreTraining. In Stage 1, the action expert is pretrained as a VA prior on vision-action pairs from a frozen VLM, bypassing the language imbalance. In Stage 2, language tokens are injected through a gated fusion mechanism that integrates VLM features while preserving the learned visuomotor prior. APT applies to mainstream VLA architectures, including the $π$ and GR00T-style architectures. Comprehensive experiments validate that APT achieves consistent gains on unseen instructions and compositional tasks. Project Page: https://xukechun.github.io/papers/APT/

We propose Ambient Diffusion Policy, a simple and principled method for imitation learning from suboptimal data in robotics. High-quality, task-specific robot data is expensive and time-consuming to collect, while suboptimal datasets with lower-quality or out-of-distribution demonstrations are abundant. Existing methods that co-train on both data sources in robotics often fail to separate the meaningful and the harmful features in the suboptimal samples. In contrast, our method extracts only the useful features by introducing a new axis to co-training in robotics: noise-dependent data usage. Ambient Diffusion Policy restricts the contribution of suboptimal data during training to only the high and low diffusion times. To rigorously justify our approach, we first observe that robot action data exhibits a spectral power law. This induces two important properties on the optimal Diffusion Policy that we exploit: a global-to-local hierarchy and locality. We theoretically formalize this discussion using a simplified model. Our experiments validate Ambient Diffusion Policy on four types of suboptimal action data (noisy trajectories, sim-to-real gap, task mismatch, and large-scale data mixtures) across six tasks. The results show that it effectively learns from arbitrary sources of suboptimal data. Notably, it outperforms existing co-training baselines by up to 33% when scaled to Open X-Embodiment - a large dataset with heterogeneous data quality and unstructured distribution shifts. Overall, Ambient Diffusion Policy increases the utility of suboptimal demonstrations and expands the set of usable data sources in robotics.

Transformers dominate modern sequence modeling, but their quadratic attention incurs substantial computational cost. Subquadratic architectures offer a scalable alternative. However, it remains unclear which designs yield the most effective sequence models. We compare three leading approaches: xLSTM, Mamba-2, and Gated DeltaNet. We evaluate these models on tasks with complex dependencies: (1) code-model pre-training, (2) distillation of code models from large language models, and (3) pre-training of time-series foundation models. Across these settings, xLSTM delivers the strongest overall performance. To explain xLSTM's advantage, we present a unified formulation and analyze the underlying architectural mechanisms, focusing on state tracking and memory dynamics. Our results show that xLSTM enables more flexible and stable memory correction via its gating scheme. We corroborate these findings on controlled synthetic length-generalization tasks. Overall, our findings indicate that xLSTM's gains on complex tasks stem from robust state tracking and accumulation.

We study multimodal learning under missing modalities, with particular motivation from bioscience applications in which heterogeneous modalities are often only partially available when decisions need to be made. We propose Latent World Recovery (LWR), a framework built on two key ideas: (i) modality-specific embeddings from different modalities are aligned in a shared latent space, and (ii) a unified representation is constructed by fusing only the embeddings of the modalities that are actually available at both training and inference time. Rather than imputing missing modalities or requiring a fixed modality set, LWR treats each modality as a partial perception of an underlying latent state and performs availability-aware representation learning directly from the observed modalities. This combination of neighbor-based latent alignment and availability-aware modality fusion enables robust multimodal prediction under partial observation, while avoiding error propagation from explicit reconstruction of missing modalities. We evaluate the proposed framework on real-world incomplete multi-omics benchmarks and demonstrate that it provides an effective approach to downstream tasks such as cancer phenotype classification and survival prediction.

Language-model post-training is the main stage at which model behavior is shaped, yet it still largely involves optimization of scalar rewards that summarize diverse desiderata. This abstraction gives practitioners little visibility into what their data actually teaches models, allowing spurious correlations to be learned by a model and inducing undesirable behaviors such as over-stylization and sycophancy. To address this problem, we ask: can we inspect a preference dataset before optimization and decide, at the level of concepts, which behaviors a model should be allowed to learn? Motivated by this, we introduce a data-centric post-training pipeline that uses interpretability protocols to develop statistical hypotheses for the latent concepts separating preferred from dispreferred generations, making them explicit for fine-grained user feedback. Building on this view, we unify several interpretability-based training protocols as ways of shaping rewards via feature or data interventions. Empirically, we show that our pipeline diagnoses undesirable signals in existing preference data, mitigates off-target learning, and can also help amplify or shape desired properties such as safeguards and model personality. More broadly, our results suggest that interpretability can turn post-training from optimizing opaque proxy rewards into a process of auditing and sculpting the learning signal itself.

Multi-robot collaboration allows robots to efficiently take on a wide range of tasks, from moving a couch through a doorway to assembling structures on a construction site. However, achieving such coordination in mobile multi-robot settings remains challenging: centralized methods conditioned on the combined observations of a team scale poorly with team size, and decentralized methods that train one policy per robot often require explicit alignment procedures or information sharing at inference time to overcome partial observability. Our key insight is that the visuomotor priors of pretrained vision-language-action (VLA) models should enable reactive, decentralized collaboration from each robot's local observations alone, without these inference-time assumptions. We propose CHORUS, a framework that adapts a single VLA backbone to control diverse, multi-robot teams. At inference time, each robot runs an independent copy of CHORUS, conditioned only on its own observations and a robot-identifying prompt. In real-world experiments including mobile tape measurement, library book handovers, and laundry basket lifting, CHORUS achieves a 64% point improvement over decentralized, from-scratch models, improves reactivity to teammate behavior by 40% points, and outperforms centralized baselines. Together, these results show that a shared VLA backbone is capable of achieving decentralized multi-robot collaboration, without per-robot policies or inter-robot communication at inference.

Hematoxylin and eosin (H&E) staining is the cornerstone of histopathology, yet scalable, quantitative analysis of H&E whole-slide images (WSIs) remains a central challenge in computational pathology. We present Atlas H&E-TME, an AI-based system built on the Atlas family of pathology foundation models that predicts tissue quality, tissue region, and cell type labels across multiple cancer types, yielding over 4,500 quantitative readouts per slide at cell-level resolution. A key challenge to validating such systems is overcoming morphological ambiguity inherent to H&E-only ground truth and the limited scalability of more informed references drawing on modalities such as immunohistochemistry (IHC). We address this with a dual validation framework combining biologically grounded depth with technical and morphological breadth. For depth, we propose an IHC-informed multi-pathologist consensus protocol that substantially improves inter-rater agreement over conventional H&E-only annotation. This yields a molecularly grounded reference against which we compare Atlas H&E-TME and pathologists working from H&E alone. For breadth, we benchmark Atlas H&E-TME on over 200,000 high-confidence H&E-only pathologist annotations across 1,500+ cases spanning eight cancer types and their most common metastatic sites, with subtypes covering >90% of clinical cases per cancer type, drawn from 25+ sources and 8+ scanner models. Benchmarked against the IHC-informed consensus, Atlas H&E-TME matches or exceeds pathologist H&E-only performance and generalizes consistently and robustly across this broad morphological and technical scope. In doing so, Atlas H&E-TME turns the H&E slide -- the most ubiquitous data in pathology -- into a scalable, quantitative window into the tumor and its microenvironment, laying a foundation for the next generation of tissue-based biomarkers in translational and clinical research.

General-purpose agents such as OpenClaw are increasingly used as autonomous tool users, but their coding ability is difficult to measure under SWE-bench: a generic agent does not by itself satisfy the clean Docker workspace, patch, and prediction contract required for scoring. We introduce Claw-SWE-Bench, a multilingual SWE-bench-style benchmark and adapter protocol that makes heterogeneous agent harnesses, or claws, comparable under fair settings including a fixed prompt, runtime budget, workspace contract, patch extraction procedure, and evaluator. The full benchmark contains 350 GitHub issue-resolution instances across 8 languages and 43 repositories, drawn from SWE-bench-Multilingual and SWE-bench-Verified-Mini after future-commit cleanup. We also release Claw-SWE-Bench Lite for faster validation, which is an 80-instance subset selected by a cost-aware, rank-aware procedure over 17 calibration columns. On the full benchmark, OpenClaw with a minimal direct-diff adapter scores only $19.1\%$ Pass@1, whereas the full adapter reaches $73.4\%$ with the same GLM 5.1 backbone, showing that adapter design is essential for enabling OpenClaw-style harnesses to perform coding tasks effectively. Across an OpenClaw $\times$ nine-model sweep and a five-claw $\times$ two-model sweep, model choice changes Pass@1 by $29.4$ pp and harness choice by $27.4$ pp under fixed models; systems with similar accuracy can differ substantially in total API cost. Claw-SWE-Bench therefore treats harness and cost accounting as first-class axes of SWE-style coding-agent evaluation, providing both a full benchmark and a low-cost reference set for reproducible comparison. The data is available at https://github.com/opensquilla/claw-swe-bench and https://huggingface.co/datasets/TokenRhythm/Claw-SWE-Bench.

Domain fine-tuning degrades the safety of large language models: fine-tuned specialists readily comply with harmful prompts framed in domain language. Existing inference-time defenses that mix logits from a safe anchor model require both models to share a vocabulary, which rules them out for the cross-family specialists where safety is most degraded. We present ALIGNBEAM, a training-free method that lifts this restriction by translating anchor logits into the target model's vocabulary token-by-token at each decoding step; a small LLM judge then selects the safest among K candidate continuations. No weights are changed, and the safety-utility trade-off can be tuned at deployment without retraining. Across both cross-vocabulary and same-vocabulary evaluation pairs, ALIGNBEAM substantially raises refusal on adversarial benchmarks while keeping task accuracy and inference overhead within practical bounds. The results show that safety alignment can be transferred between model families at inference time, without touching either model's weights.

Memory is not merely the storage of data; it is the scaffolding of reality. When biological memory fades, the world does not simply turn black; it regresses into an unrecognizable chaos. Echoes of the Prior is an interactive installation that attempts to visualize this subjective phenomenology of forgetting. By inducing controlled synaptic decay within a Feed-Forward 3D Reconstruction model, we create an artistic analogy for the erosion of the brain's predictive priors. We position the Neural Network not as a tool for engineering, but as a cognitive proxy - a silicon brain whose structural degeneration evokes the disorienting, poetic, and terrifying experience of losing one's grip on the world. Ultimately, we offer this framework as a catalyst, inviting the wider community to explore the uncharted potential of neuromorphic aesthetics in visualizing the fragility of intelligence. Interactive demo see https://decart-4d.github.io/.

Sound source distance estimation (SDE) is a critical capability in human-robot interaction. An inappropriate interaction distance not only reduces the reliability of speech acquisition and understanding, but also compromises the naturalness and comfort of the interaction. Most existing SDE methods rely on microphone arrays, however, multi-microphone systems typically require careful hardware synchronization, geometric calibration, and additional space and computational resources, which limits applicability to size-constrained and computability-limited embodied platforms. To alleviate these issues, we propose Fast-SDE, a lightweight single-microphone SDE framework that is suited for deployment on robot platforms with limited computational resources and strict size constraints. Specifically, Fast-SDE employs a subband-based backbone that decomposes the frequency axis into multiple subbands, rather than processing the entire spectrum with a wide full-band backbone. A shared subband encoder then maps each subband to a compact latent representation and learns the relationship between acoustic structure and time-frequency patterns. Finally, a lightweight regression head converts the fused subband representations into the estimated distance. Extensive simulation and real-world experiments demonstrate the merits of the proposed method. To benefit the broader research community, we have open-sourced our code at https://github.com/JiangWAV/FAST-SDE.

Inverse problems governed by partial differential equations (PDEs) are central to computational mechanics and are commonly solved by adjoint-based optimization, while physics-informed neural networks (PINNs) have emerged as a flexible alternative. Their relative performance remains difficult to assess because the two approaches are often compared under different formulations, parameterizations, optimizers, and regularization choices. We present a fair comparison of adjoint optimization and PINNs for PDE-constrained inverse problems. From a common abstract formulation, we instantiate both methods on identical domains, governing equations, observation models, and regularization terms, while matching the optimizer, unknown parameterization, and arithmetic precision wherever applicable. The benchmarks include unsteady Burgers, noisy Darcy permeability inversion, three-dimensional Allen--Cahn reaction identification, and unsteady Navier--Stokes viscosity identification. The results show that the representation of the unknown largely determines the preferred method: grid-based fields favor the discrete adjoint, whereas neural representations are native to PINNs and relevant for closure and constitutive modeling. For time-dependent problems, adjoint inversion can be dominated by trajectory storage and differentiation, while PINNs provide satisfactory reconstructions at lower cost. A PINN-warm-started adjoint strategy then recovers adjoint-level accuracy at substantially reduced cost.

High-precision robotic manipulation requires fine-grained spatial reasoning that is often difficult to achieve with RGB-only policies due to depth ambiguity and perspective scale issues. Policies that leverage 3D information directly, such as those based on point clouds, offer a stronger geometric prior over purely image-based ones, yet their performance remains highly task-dependent. We hypothesize that this discrepancy may be due to the spectral bias of neural networks towards learning low frequency functions, which especially affects architectures conditioned on slow-moving Cartesian features. We thus propose to map point clouds from Cartesian space into high-dimensional Fourier space, effectively equipping the point cloud encoder with direct access to high-frequency features. We experimentally validate the use of Fourier features on challenging manipulation tasks from the RoboCasa and ManiSkill3 benchmarks and on a real robot setup. Despite their simplicity, we find that Fourier features provide significant benefits across diverse encoder architectures and benchmarks and are robust across hyperparameters. Our results indicate that Fourier features let policies leverage geometric details more effectively than Cartesian features, showing their potential as a general-purpose tool for point cloud-based imitation learning. We provide source code and videos on our project page: https://fourier-il.github.io/fourier-il

AI coding assistants now support a growing share of software work, from quick scripts to production applications. Yet these agents remain largely stateless: each new session re-reads project files, re-derives prior decisions, and - most costly - may repeat debugging attempts that already failed. Reconstructing this context can consume an estimated 5,000-20,000 tokens per session; the bottleneck is often not model capability but missing project memory. We present projectmem, an open-source, local-first memory and judgment layer for AI coding agents. projectmem records development as an append-only, plain-text event log of typed events - issues, attempts, fixes, decisions, and notes - and deterministically projects that log into compact, AI-readable summaries served through the Model Context Protocol (MCP). Beyond storage, projectmem adds a deterministic pre-action gate that warns an agent before it repeats a previously failed fix or edits a known-fragile file. We frame this as Memory-as-Governance: memory that does not merely answer the agent but acts on its next action. The system runs fully offline with no telemetry; its immutable log also serves as a provenance trail for reproducible, auditable AI-assisted development. projectmem ships as a three-dependency Python package (14 MCP tools, 19 CLI commands, 37 automated tests) and is evaluated through a two-month self-study across 10 projects comprising 207 logged events. Source code: https://github.com/riponcm/projectmem.

Enterprise security was built to govern data boundaries: the protected surface was data at rest and in transit, and the controls -- access control, data-loss prevention, perimeter inspection -- governed crossings of that boundary. Production AI agents dissolve this assumption. An agent reads context, calls tools, invokes connectors, and modifies systems of record on an enterprise's behalf, so risk moves inside the workflow, into sequences of individually-permitted actions that may transform a business process no one authorized. Existing policy engines do not extend to this regime: they evaluate request-time decisions against atomic principals, where agentic systems require stateful evaluation against composite principals whose authority attenuates through delegation chains. We present a reference architecture for the runtime governance of production agents, built from four composable primitives: a five-plane decomposition (a reasoning plane that adjudicates intent, and four enforcement planes -- network, identity, endpoint, data -- that realize the decision), stop-anywhere mediation, composite principals with capability attenuation, and audit as a structured evidence substrate. We define a taxonomy of six interruption primitives that generalize allow and deny, state and argue for four correctness invariants, and demonstrate the foreclosure of seven production-agent threats across five concrete workflows. A reference implementation of the policy-engine core supplies measured evidence: attenuation correctness and evidence reconstructability hold on every trial, adjudication runs in single-digit microseconds, and the audit substrate's tamper-evidence behaves exactly as designed. We are explicit about scope: the architecture governs delegated action, not model behavior, and a full-system evaluation against a live agent benchmark is the invited next step.

Segmenting the Circle of Willis (CoW) from Magnetic Resonance Angiography (MRA) is challenging due to complex topology and thin vascular structures that are prone to fragmentation. Standard Convolutional Neural Networks (CNNs) often fail to capture these topological constraints, resulting in "broken vessel" artifacts. To address this, we propose the Anatomically Conditioned Recurrent Refinement U-Net (AC2RUNet). Our architecture decouples segmentation into two streams: a Static Stream that extracts invariant anatomical features and a lightweight Dynamic Stream that iteratively refines topological errors over time. We further introduce a dynamic curriculum learning strategy that transitions from high-recall geometric supervision to topology-aware constraints. Validated on the TopCoW dataset, AC2RUNet substantially reduces Hausdorff Distance (4.72 mm vs 9.17 mm) and Betti number errors (0.19 vs 0.40), improving topological connectivity over the nnU-Net baseline while maintaining comparable volumetric Dice.

ARC tests in-context rule induction: given a few input-output demonstrations, a model must infer the hidden rule and apply it to a new query. While many approaches express ARC rules through language, code, or symbolic programs, ARC itself is visual-symbolic: rules appear as grid transitions over objects, colors, shapes, and spatial relations. We introduce Loop-OWM, an object-centric world-modeling architecture that learns these rules as composable transitions over structured states. It combines color-prototype slots, demonstration-conditioned task summaries, and a looped transition model with dense propagation and slot-conditioned correction. On both ARC-1 and ARC-2, Loop-OWM outperforms non-looped and looped baselines with comparable or fewer parameters. These results suggest that ARC rules can be learned not only as language descriptions or searched programs, but also as transitions over visual-symbolic world states.

Laser-induced forward transfer (LIFT) is a nozzle-free laser-assisted printing method that provides an advanced manufacturing route for spatially selective deposition of functional inks, nanoparticle suspensions, polymers, hydrogels, biological materials, and other difficult-to-nozzle formulations. The apparent simplicity of LIFT, however, conceals a strongly coupled laser-liquid interaction. Laser energy is absorbed within a confined donor architecture, converted into thermal and plasma responses, and then transformed into bubble-mediated motion of the donor material. The cavitation bubble provides the transient mechanical bridge between optical energy deposition and the hydrodynamic ejection process. This chapter presents LIFT from a multiscale perspective centered on bubble dynamics and material ejection. It first reviews major LIFT donor architectures. Then, it examines how donor ribbon design, absorbing-layer properties, laser parameters, material rheology, control bubble inception/growth, jet formation, droplet breakup, and final deposition. Modeling approaches are discussed as tools for connecting experimental observations across time and length scales, ranging from reduced-order estimates to interface-resolving simulations and data-driven process maps. As one illustrative mechanistic example, thermal-only, plasma-mediated, and coupled plasma-thermal-thermoelastic frameworks for early-stage bubble inception are briefly compared to show how different inception assumptions can provide initial conditions for downstream bubble growth and jetting models. This chapter concludes by identifying opportunities for bubble-aware donor design, time-resolved diagnostics, benchmark datasets, and predictive LIFT process maps based on intermediate bubble and jet observables.

The realization of on-chip photonic circuits requires scalable and deterministic single-photon emitters (SPEs) at room temperature, which remain a challenge in van der Waals materials. In this work, we report a novel three-step fabrication process for the generation of spatially controlled SPE arrays in hexagonal boron nitride (hBN). The process comprises site-selective gallium (Ga) focused ion beam milling, nanoscale conformal carbon deposition over the patterned regions, and subsequent thermal annealing. The synergistic combination of these steps resulted in a site-correlated emitter yield of ($\sim 89\%$) across 100 fabrication sites. Second-order autocorrelation measurements revealed pronounced three-level emitter dynamics where the best emitters exhibited high purity ($g^{(2)}(0)=0.15 \pm 0.09$).To the best of our knowledge, this is the first lithography-free, direct-write approach combining Ga-ion milling, selective carbon engineering, and thermal annealing to deterministically generate \hBN{} \SPE{}s. The reproducibility of the method is validated across multiple independently fabricated samples. These results establish a scalable, lithography-free pathway toward on-demand SPE arrays relevant to integrated quantum photonics.

Multimodal image fusion aims to integrate complementary information from different modalities into a fused image that preserves rich local details while maintaining globally consistent appearance. Existing approaches build shared representations on 2D feature grids, which excel at modeling local structures but offer limited leverage over image-level global appearance factors. To balance these objectives, we introduce a compact 1D token interface based on a frozen pretrained image tokenizer for modeling non-local appearance/base factors. Rather than using the tokenizer as a reconstruction backbone, our design uses the 1D token space as a global carrier while retaining the 2D spatial pathway for local structure restoration. Specifically, we introduce Selective Token Editing (STE), which sparsely updates/replaces a small set of critical tokens, providing a lightweight mechanism to steer global appearance coherence while keeping the fusion backbone unchanged and avoiding extra losses. Experiments on four commonly used benchmarks show that our method achieves the best overall performance, with consistent, multi-metric improvements in both global coherence and local fidelity. Project page: https://zju-xyc.github.io/1D-Fusion-Project-Page/

Temporal grounding--returning the interval $[t_s, t_e]$ for a natural-language query over a video--is the language interface to long-form video, yet has been studied on short videos; the dynamics of hour-scale natural-language grounding remain underexplored. We take the position that at hour-scale, the binding constraint is search, not recognition: Video-LLMs are bottlenecked not by localizing a nearby event, but--given a natural-language query--by searching for the relevant region of a long video. To test this, we release ExtremeWhenBench, the first open hour-scale grounding benchmark (2,273 queries over 194 videos, mean 75.7 min, max 9 hr) with an open-form query distribution. Every open Video-LLM collapses while a frame-level retrieval baseline outperforms them; a failure taxonomy attributes 85% of failures to search; and a retrieve-then-ground hybrid recovers 6.7x over the monolithic Video-LLM--mirroring retrieve-then-read in open-domain QA.

Vision-Language-Action (VLA) models provide a natural language interface to robot control, but the mapping from language to behavior is often brittle and unintuitive: semantically similar instructions can induce drastically different behaviors, while some capabilities may not be elicitable through prompting alone. As a result, both human instructions and zero-shot language models can fail to reliably steer VLAs toward successful task execution. In this work, we propose a framework that interactively searches for language sequences that improve closed-loop VLA task performance, distills these sequences into a test-time language feedback policy (LFP), and learns an improvement head that predicts when language steering will improve performance. We conformalize this improvement head to prevent harmful steering interventions, where the LFP decreases task performance relative to the original instruction on out-of-distribution scenarios. Crucially, our approach operates on arbitrary frozen pre-trained VLAs, requiring neither access to the original training distribution nor fine-tuning of the underlying model. On seen environments, our conformalized LFP improves base VLA performance by 24.7% in simulation and 65.0% in hardware. On visual and semantic perturbations, our conformalized LFP has strong harmlessness guarantees, and produces recovery behaviors not observed with open-loop prompting.

Large language models (LLMs) now reach expert-level scores on medical licensing exams, encouraging the assumption that high scores imply safe medical judgment while patients increasingly use them for health advice. We show this assumption is fragile: when misleading context is injected into questions that LLMs originally answer correctly, they abandon the correct answer. We call the ability to maintain correct judgment under adversarial context epistemic resilience, and introduce MedMisBench to measure it. MedMisBench contains 10,932 medical question items and 48,889 misleading context-option pairs spanning medical reasoning, agentic capability, and patient-journey evaluation. Across 11 model configurations, mean accuracy falls from 71.1% on original questions to 38.0% under focused misleading context, with 51.5% attack success. The most damaging injections are formal, rule-like fabrications: authority-framed falsehoods reach 69.5% attack success and exception-poisoning claims reach 64.1%. A 14-member clinical panel from 7 countries identified serious potential harm in 38.2% of reviewed cases. MedMisBench exposes a structural blind spot in LLM evaluation in medical settings: existing benchmarks measure what models know, but not whether they preserve correct medical judgment under misleading context.

As Artificial Intelligence models grow in complexity, interpretability has become an indispensable tool for understanding, debugging, and controlling their computations. However, interpretability lacks general theories to deductively design interpretable methods. This gap between theories and methods results in a fragmented literature and inconsistent evaluation protocols. To fill this gap, we introduce the Standard Interpretable Model (SIM), a general theory grounded in Lagrangian mechanics that enables the deductive design of interpretable methods. Specifically, the SIM summarises, in a set of premises, what interpretability is for a target user. From these premises, the SIM systematically derives interpretability symmetries and corresponding constraints, which shape the landscape of a Lagrangian whose minima correspond to optimal interpretable models. To reach the minima, one can either update the parameter values of an opaque model to make it more interpretable or compile constraints into an interpretable architecture. We empirically show that the SIM identifies and solves limitations of existing methods (including traditional, concept-based, and mechanistic interpretability), highlights underexplored research directions, and informs the design of core programming interfaces. Beyond being a research method, the deductive nature of the SIM offers pedagogical grounding for interpretability curricula and may shift the scientific community's perspective of a discipline that has long been fragmented.

In Decentralized Training and Decentralized Execution (DTDE) for cooperative Multi-Agent Reinforcement Learning (MARL), action-advising-based knowledge sharing promotes interpretable and scalable cooperation among agents. However, current action advising approaches often adhere too much to the teacher's guidance without evaluating teacher-student compatibility, which causes excessive advising, suboptimal stability, and degraded performance. To overcome these challenges, this paper presents a Consensus-based Communication and Knowledge Sharing (CCKS) framework, which allows agents to adopt recommendations based on consensus-derived constraints and to follow the teacher's instructions more smartly. This mechanism enables agents to balance exploration and learning from experienced teachers, improving overall performance. The key is the consensus model construction, for which we propose to employ contrastive learning to construct consensus models based on local observations in the agents' training phase. In action selection, agents score and choose actions based on consensus and shared knowledge. Designed as a plug-and-play solution, CCKS integrates seamlessly with existing DTDE algorithms. Experiments conducted in the Google Research Football environment and the complex StarCraft II Multi-Agent Challenge demonstrate that the integration with CCKS significantly improves cooperation efficiency, learning speed, and overall performance compared with current DTDE baselines. The code is available at https://github.com/yuanxpy/CCKS.

Post-training quantization lets large text-to-image diffusion transformers run on consumer GPUs, yet the hardware-specific trade-offs are seldom measured directly. We quantize Ideogram 4.0 - a 9.3B flow-matching diffusion transformer (DiT), shipped as two separate-weight copies of a single-stream 34-layer backbone for classifier-free guidance and conditioned by a Qwen3-VL-8B encoder - for Ampere RTX 3090 GPUs, which lack FP8 tensor cores. Our INT8 W8A8 recipe (per-channel weights, per-token dynamic activations, SmoothQuant, and mixed-precision protection of a small high-fragility layer set) holds the FP8 quality ceiling: on a 200-prompt benchmark the paired same-seed bootstrap CI for INT8-FP8 includes zero on both Pick and CLIP, while INT8 improves on NF4 by $+1.9$ CLIP (95% CI $[+1.21,+2.64]$, excluding zero). A per-category OCR analysis, to our knowledge unreported for this model class, confirms text legibility is preserved, and an ablation isolates protection of the FFN down-projections as the dominant quality lever. Our GGUF Q4_K quantization beats NF4 at equal on-disk size and is the Pareto winner on the quality-memory frontier, with paired confidence intervals excluding zero (Q8_0 is quality neutral). Finally, we characterize where 8-bit quantization helps and where it does not: INT8's weights match FP8's footprint rather than shrink it, so a speed gain on Ampere awaits a fused INT8 kernel.

Neural network pruning reduces model size by removing less important parameters while aiming to preserve predictive performance. Although the Lottery Ticket Hypothesis (LTH) shows that sparse subnetworks can match dense networks when trained from suitable initializations, its iterative pruning procedure requires multiple complete training cycles. This work evaluates progressive magnitude-based pruning as a single-cycle alternative. The method gradually increases sparsity during training using a linear schedule and updates pruning masks based on active weight magnitudes. We conduct systematic experiments on CIFAR-10 and MNIST across ResNet, VGG-style, and LeNet architectures, comparing the proposed method with representative iterative and initialization-based pruning baselines, including LTH, SNIP, and GraSP. On CIFAR-10, the method achieves 95.12\% accuracy on ResNet-18 at 72.9\% sparsity, compared with 90.5\% reported for LTH. At extreme sparsity, it achieves 93.13\% accuracy on a VGG-like architecture at 97\% sparsity, compared with approximately 92.0\% for SNIP, and 93.44\% accuracy on VGG-19 at 97.97\% sparsity, compared with 92.19\% for GraSP at 98\% sparsity. A sparsity-accuracy analysis on ResNet-18 further shows that accuracy remains within 0.1 percentage points of the dense baseline across 70--85\% sparsity. These results indicate that progressive magnitude-based pruning provides an effective single-cycle approach for neural network sparsification under the evaluated settings.

Diffusion large language models (dLLMs) offer an efficient alternative to autoregressive models through parallel decoding, yet existing post-training methods largely rely on random masking strategies that overlook intrinsic token dependencies. In this work, we present an empirical analysis of attention in dLLMs and show that tokens attending more strongly to unmasked context exhibit greater generation stability and play a critical role in reasoning. Motivated by these findings, we propose AGDO, an attention-guided denoising and optimization framework that aligns both training and optimization with attention-derived dependencies. AGDO determines the denoising order based on attention structure and emphasizes attention-critical tokens during supervised fine-tuning and reinforcement learning. Experiments on mathematical and coding benchmarks demonstrate that AGDO consistently improves reasoning performance, outperforming state-of-the-art post-training methods for dLLMs.

Advanced AI systems have extensive knowledge of their environments; in fact, their knowledge may (far) exceed that of their developers or users. Consequently, a desirable property for an AI system is that it is honest -- that it accurately reports its beliefs about the world. Designing an AI system to be honest may be difficult, especially if we want to ask it questions about latent variables in the environment -- variables which are hidden from the human interacting with it. This gives rise to the problem of eliciting latent knowledge (ELK): the problem of training an AI agent to honestly report its beliefs. In this paper, we make ELK formally precise using Causal Influence Diagrams (CIDs). CIDs can be used to describe the relationship between an agent's training environment and its subjective representation of the world. We use CIDs to formalise the distinction between observable and latent variables, to specify what exactly it means for an agent to be honest, and to formally define goal misgeneralisation. We show that, under certain circumstances, developers can incentivise an agent to honestly answer questions by providing correct feedback during training. However, a natural, but undesirable, way for an agent to generalise is to provide answers which humans would evaluate as true, rather than honest answers. We prove an impossibility theorem stating: There is no feedback-based training strategy that depends only on agent behaviour and with certainty produces an honest agent, even if feedback is perfect during training.

Cross-domain day-night re-identification (ReID) is fundamentally challenged by the substantial visual appearance discrepancies between daytime and nighttime scenes. Existing fully supervised methods rely heavily on labor-intensive annotations, which are costly and exhibit limited generalization across domains. In this work, we investigate unsupervised day-night ReID and propose a novel framework that synergistically combines prompt learning and prototype-based representation learning to associate identities across domains without requiring manual labels. Our approach follows a progressive two-stage training strategy. In the first stage, we exploit the vision-language model to generate instance-specific textual prompts in an annotation-free manner. We employ an instance-level alignment mechanism to embed visual features and textual prompts into a unified semantic space, aligning unlabeled day/night images with learnable prompts via instance-aware dynamic-bias adaptation. In the second stage, we construct domain-specific prototype memory banks and introduce two complementary modules: i) an intra-domain identity association module to enhance feature discriminability within each domain, and ii) a cross-domain prototype matching module to reliably identify positive and negative prototype pairs, thereby establishing robust identity correspondences across day and night. Extensive experiments on public benchmarks validate the effectiveness of our method. Under the unsupervised setting, our framework attains Rank-1 accuracy comparable to state-of-the-art fully supervised methods.

Large language models (LLMs) in medicine are mainly evaluated using multiple-choice question answering (MCQA), which can overestimate real clinical ability due to guessing strategies and answer biases. To address these limitations, we introduce an expanded and more challenging benchmark based on Polish medical exams, adding over 15,000 questions, two new domains, and four structural modifications that reduce MCQA-specific artifacts and better test reasoning. We evaluate 21 LLMs and show that evaluation design strongly affects results. Under our harder setup, the best model (Qwen3.5-122B) drops by 28.4 and 31 pp on English and Polish exams, respectively. Despite low evidence of data contamination, standard MCQA scores do not reliably reflect true medical competence. To facilitate further research, we make our benchmark publicly available.

Decision-relevant building damage assessment is critical for prioritizing resources and recovery after a disaster, yet most automated methods either flatten damage into a single severity scale (no damage, minor, major, destroyed) or require paired pre- and post-event imagery that is often unavailable for emerging hazards. This paper presents Damage-TriageFormer, a single-image, post-event, footprint-conditioned model that produces a damage typology rather than a severity scale. We contribute: (1) DamageTriage-Bench, a new benchmark built from NOAA Emergency Response Imagery across Hurricane Michael (2018), Hurricane Helene (2024), and the 2025 Los Angeles wildfire complex, with five typology classes that distinguish roof damage from structural damage and, within each, partial from total extent; and (2) Damage-TriageFormer, which extends a DINOv3 ViT-L backbone with a Simple Feature Pyramid for higher-resolution instance pooling, a two-stage gated damage head, and an auxiliary severity-regression objective. Our model achieves macro F1 of 0.624 on validation and 0.619 on a held-out stratified test set, performing strongest where operational triage needs it most, with per-class F1 of 0.91 and 0.84 on undamaged buildings and total structural collapse, respectively. While the rare Total Roof Damage class remains difficult due to its limited examples and an inherently ambiguous label boundary, our results show that single-image post-event imagery can support actionable building damage typing, enabling targeted emergency response and resource allocation without a pre-event reference.

Cold-start item recommendation remains a persistent challenge in real-world systems due to the absence of interaction histories. While prior models attempt to bridge this gap using item content features, they universally suffer from the \textbf{seesaw dilemma}: enhancing performance for cold items inevitably degrades performance for warm items, and vice versa. We identify that this dilemma stems from a fundamental \textbf{distributional disparity}: warm item embeddings occupy a complex ``behavioral manifold" shaped by rich interaction signals, whereas cold item embeddings are constrained to a ``semantic manifold" derived solely from auxiliary content. Existing methods often force a rigid mapping between these inconsistent spaces, causing the model to sacrifice the precision of warm representations to accommodate cold ones. To address this, we propose \textbf{DiffCold}, a diffusion-based generative model that unifies warm and cold representations. Unlike GANs or VAEs, DiffCold leverages conditional diffusion to reconstruct warm item embeddings from content, preserving the underlying manifold structure without degradation. We further tailor this paradigm with two specific designs: a \textbf{Retrieval-enhanced Aggregator} that initializes generation using semantically similar warm items to bypass inefficient noise, and a \textbf{Simulation-based Representation Alignment} module that enforces distribution consistency between generated and real embeddings via contrastive learning. Experiments on three benchmarks confirm that DiffCold resolves the seesaw dilemma, consistently outperforming state-of-the-art methods across all metrics.

Multivariate time-series data often exhibit complex temporal dependencies, irregular sampling, and heterogeneous dynamics across multiple time scales, making accurate sequence modeling particularly challenging. Traditional recurrent neural networks (RNNs), such as Long Short-Term Memory (LSTM) networks, operate in discrete time and may struggle to effectively capture continuous and irregular temporal behaviors. Liquid Neural Networks (LNNs) address some of these limitations through continuous-time dynamics, but standard LNN architectures typically rely on a single dynamical system, limiting their ability to model heterogeneous temporal patterns. To address these challenges, we propose a Multi-Rate Mixture-of-Experts (MR-MoE) framework built on top of Liquid Neural Networks. In the proposed architecture, multiple LNN-based experts operate at distinct time scales, enabling the model to explicitly separate fast-changing dynamics from slow-evolving temporal trends. A gating network further enables adaptive expert specialization based on input conditions. In addition, we incorporate both feature-level and temporal attention mechanisms to improve robustness, interpretability, and long-range dependency modeling. Feature-level attention suppresses noisy or irrelevant variables, while temporal attention selectively focuses on informative historical states. We evaluate the proposed framework on a complex multivariate time-series prediction task and compare it against strong baselines, including LSTM, monolithic LNN, and standard MoE models. Experimental results demonstrate that the proposed MR-MoE framework consistently achieves improved AUROC and AUPRC performance while maintaining favorable computational efficiency. These results highlight the effectiveness of combining continuous-time dynamics, multi-scale expert decomposition, and adaptive attention mechanisms for time-series modeling.

Many autonomous driving systems are increasingly incorporating foundation models to improve generalization and handle long-tail scenarios. However, this trend introduces two key challenges: (i) the manual and labor-intensive process of designing and integrating new models, and (ii) the lack of intelligent, dynamic scheduling mechanisms to meet strict real-time constraints. While Large Language Model (LLM)-based agents offer a promising avenue for automation, existing frameworks are ill-suited for autonomous driving. Specifically, they fail to distinguish between the fundamentally different requirements of system design and real-time scheduling, treat modules as opaque black boxes, and are not designed for continuous operation. To address these limitations, we propose DrivingAgent, a novel agent framework tailored to the dual challenges of autonomous driving system design and scheduling. In the design phase, DrivingAgent automates module development by interpreting system architecture, generating code, and validating modules via super-network training. In the scheduling phase, it employs a lightweight LLM trained with reinforcement learning to dynamically orchestrate system modules in real time, supported by a structured memory that integrates long-term storage with timestamped short-term context. Experimental results demonstrate that DrivingAgent achieves a superior speed--accuracy trade-off on both the nuScenes and Bench2Drive benchmarks.

Controlling the output of Large Language Models (LLMs) is a central challenge for their reliable deployment, yet a clear understanding of the involved trade-offs remains elusive. Current approaches to conditioning are often evaluated with a narrow focus on their effectiveness at injecting or removing a target concept, neglecting generation quality. We systematically investigate a range of conditioning methods in both injection and removal scenarios. We find that efficient steering methods frequently achieve conditioning at a steep cost to fluency. Furthermore, we identify a critical yet previously overlooked interaction with the training paradigm: activation steering methods are far less effective on instruction-tuned models than on their base counterparts. Simple prompting and full-fledged supervised fine-tuning, on the other hand, are viable options for concept injection, but are not as good at concept removal. Finally, cheaply computed textual metrics highly correlate to costly LLM-as-judge scores, and provide insights on the behavior of conditioning methods.

The Stokes parameters are three real parameters that completely characterize partially coherent optical fields spanned by two modes -- whether a pair of polarization or spatial modes -- and their use is thus ubiquitous in optics. Because the Stokes parameters are defined through an expansion of the $2\times2$ coherence matrix in terms of the Pauli matrices, they cannot be applied to optical fields comprising three modes, which are described by a $3\times3$ coherence matrix. Examples of such fields include the polarization of non-paraxial fields (spanned by three orthogonal polarization modes), and fields comprising three spatial or temporal modes. It has long been theorized that the $3\times3$ Gell-Mann matrices -- developed in high-energy particle physics -- can serve as a basis for $3\times3$ optical coherence matrices, with 8~expansion coefficients known as the Stokes-Gell-Mann (SGM) parameters, but the measurement procedure is daunting, and the SGM parameters have not been measured directly to date in optics. Here we present the first measurements of the SGM parameters for partially coherent three-mode light in a photonic integrated platform comprising a hexagonal mesh of Mach-Zehnder interferometers. Measuring the SGM parameters on chip, from which we reconstruct the $3\times3$ coherence matrix facilitates exploring the full space of iso-entropy fields that can be inter-converted into each other unitarily, and those that share the same value of entropy and yet cannot be inter-converted unitarily. These results pave the way to utilizing multimode partially coherent light in applications involving optical communications, sensing, and information processing.

Masked diffusion language models (dLLMs) have recently emerged as a competitive alternative to autoregressive language models, with the promise of faster inference via parallel token generation. A notable limitation of the masked formulation, however, is that once a token has been unmasked it can no longer be revised, leaving dLLMs vulnerable to early sampling mistakes. To address this, a growing body of work has sought to extend masked dLLMs with self-correcting (remasking) capabilities. One appealing subset of these methods does so in a training-free, post-hoc manner based on token confidences, with encouraging early reported results. In this work, we revisit the empirical evaluation of a representative post-hoc remasking method, WINO [Hong et al., 2026], and find that under standard decoding settings (shorter block lengths) it brings little-to-no benefit over confidence-based unmasking alone [Wu et al., 2025]. Extending the evaluation to non-greedy decoding, we find that while confidence-based remasking can mitigate errors introduced by increased stochasticity to some extent, it also exacerbates the diversity collapse previously reported for confidence-based unmasking. Overall, our results show that the benefits of post-hoc confidence-based remasking are highly setting-dependent, underscoring the need for a more comprehensive evaluation framework.

While deep learning has significantly advanced image reconstruction of Electrical Capacitance Tomography (ECT), most data-driven methods map directly between capacitance and permittivity distribution, treating the sensor as a black box. This overlooks the electric potential field -- the fundamental physical link governing the nonlinear and ill-posed ``soft-field'' effect. To address this, we propose an electric potential-augmented ECT benchmark dataset designed to explicitly integrate latent physics behind ECT into the learning process. Generated via a COMSOL-MATLAB pipeline for an eight-electrode sensor as an example, the dataset comprises 20,000 randomized samples across four typical flow patterns. Crucially, alongside the conventional capacitance vectors and permittivity distributions depicted as images, each sample preserves eight excitation-wise full-field potential maps. Beyond data release, we provide illustrative evaluation protocols for both forward and inverse problems of ECT. Through comprehensive testing on both in-distribution (IID) and out-of-distribution (OOD) scenarios, we systematically demonstrate how the inclusion of electric potential maps enhances modeling accuracy and robustness. Fundamentally, the explicit inclusion of latent field information significantly lowers the barrier to integrating physical laws into ECT modeling, thereby establishing a standardized foundation for future physics-guided machine learning of ECT image reconstruction.

World Action Models (WAMs) offer a promising route for robot manipulation by using video generation models to model future scene evolution before producing control actions. However, our empirical observations reveal a phenomenon: generating plausible visual futures does not always guarantee the extraction of accurate actions. To diagnose this failure, we conduct action-head attention analysis and causal interventions. We find that the action decoder fails to focus on task-relevant interaction regions and remains sensitive to perturbations in task-irrelevant areas. This reveals a representation mismatch: hidden states optimized for visual reconstruction are not inherently organized in a form useful for low-level action control. In this paper, we propose AGRA, an Action-Grounded Representation Alignment objective that regularizes the world-action interface by aligning intermediate video diffusion features with spatially coherent semantic representations from a foundation visual encoder. We evaluate AGRA on real-world manipulation tasks. Experiments show that AGRA makes world model representations more action-grounded: by focusing the action decoder on the correct interaction regions, it improves object localization accuracy and affordance understanding, and makes the policy more robust to perturbations in task-irrelevant regions. As a result, AGRA consistently improves both in-distribution performance and out-of-distribution generalization over the baseline world action model.

Embodied intelligence now spans navigation, household assistance, manipulation, autonomous driving, aerial agents, and multimodal large-model control. This expansion has made benchmark construction a central bottleneck for reliable evaluation. Unlike static datasets, embodied benchmarks combine task specifications, environments, robot data, demonstrations, annotations, metrics, evaluation scripts, and release policies into a single evaluation system. This survey reviews the literature through a five-stage construction pipeline: requirement and task construction, data acquisition, data cleaning and annotation, benchmark suite generation and metric definition, and evaluation execution with diagnostic feedback. For each stage, the survey analyzes the transition from manual curation to traditional automation, foundation-model assistance, and agentic closed-loop workflows. It also compares qualitative construction costs across human labor, data and asset acquisition, compute and simulation, validation and debugging, governance and maintenance, and rework risk. The main conclusion is that automation does not simply reduce benchmark cost. Instead, it often shifts cost toward validation, auditability, version control, and long-term governance. Progress in embodied evaluation will therefore depend not only on larger benchmark suites, but also on construction pipelines that are diagnosable, auditable, and responsibly refreshable.

Recent progress in foundation models has shifted toward agentic behavior involving multi-step reasoning and tool use. However, open-source efforts largely focus on text-dominant settings, leaving long-horizon multimodal tasks underexplored. This gap is evident in video tasks requiring sustained temporal understanding and iterative interaction. We present InternVideo3, a framework enhancing these capabilities via Multimodal Contextual Reasoning (MCR). MCR treats understanding as a closed-loop process over a shared, evolving context containing observations, instructions, reasoning, tool actions, and memory. This frames long-video understanding as evidence accumulation and verification. To ensure efficiency, we introduce Multimodal Multi-head Latent Attention (M^2LA), a token-preserving reparameterization compressing KV-cache states while retaining the full token stream. Our staged training includes continued pretraining, short-to-long supervised fine-tuning, rule-based reinforcement learning, and on-policy distillation. Experiments show InternVideo3 achieves strong performance on benchmarks like Video-MME, MLVU, and EgoSchema. We further instantiate the model as a video agent with retrieval tools, demonstrating robust evidence-grounded behavior. Our results suggest that efficient context handling and closed-loop reasoning are vital for adapting open multimodal models toward long-horizon visually grounded agency.

Environments serve as interactive systems for large language model (LLM) based agents across diverse scenarios and play a crucial role in driving the continual evolution of model capabilities. Despite this importance, existing work lacks a systematic categorization and deep analysis. This paper systematically studies current researches on agentic environments from the perspective of the environment engineering lifecycle, covering their modeling, synthesis, evaluation and application. Specifically, the paper first introduces representative environments from the perspectives of eight attributes and eight domains, providing detailed analyses of their development paths and highlighting their core capabilities. Second, for automated environment synthesis, two paradigms are introduced, such as symbolic synthesis and neural synthesis. This paper also shows different environment evaluation methods in each paradigm. Thirdly, the corresponding environment applications from the perspective of agent-environment co-evolution are discussed. In specific, the paper characterizes the primary pathways for agent evolution in dynamic environments from four complementary perspectives: memory-centric experience evolution, orchestration-centric workflow evolution, trajectory-centric offline evolution, and exploration-centric online evolution. And three paradigms of environment evolution are identified, namely neural-driven, difficulty-driven, and scaling-driven approaches. At last, several promising future directions are discussed, including Environment-as-a-Service, Multi-agent Environments, and Neural-Symbolic Environments.

We address 4D reconstruction from partial point cloud sequences, where depth-sensor observations are incomplete, unordered, and lack explicit temporal correspondences. This geometry-only setting is challenging due to missing observations and ambiguous dynamics. While recent progress has largely relied on image-based methods, existing point-based approaches typically focus on single objects, assume relatively complete inputs, or require explicit correspondences. To address these limitations, we propose DynaTok, a point-based framework for correspondence-free 4D reconstruction from partial point cloud sequences without images. DynaTok encodes frames into compact latent tokens, aggregates incomplete observations over time with a Transformer-based spatiotemporal encoder, and decouples geometry and motion through residual tokens in a unified model. A flow-matching decoder then reconstructs complete, temporally consistent 4D point-cloud sequences conditioned on the latent tokens. Experiments on object- and scene-level benchmarks demonstrate improved reconstruction quality and temporal coherence from partial point cloud observations. Project page: https://wrchen530.github.io/dynatok/.

Randomness amplification, the task of extracting uniform private bits from biased seeds that may be partly known by a malicious third party, is of central importance in cryptography. The highest security in this task is provided by a class of quantum protocols known as device-independent, which however are challenging to integrate into scalable devices. Semi-device-independent (SDI) protocols are a promising alternative that guarantees security under few natural assumptions, such as bounds on the amount of energy used by the devices. Here, we provide the first demonstration of SDI randomness amplification on an integrated silicon photonic chip, achieving a throughput rate of 20 Mbps suitable for practical applications. This rate is achieved through a novel technique for SDI entropy certification, which delivers strictly tighter von Neumann entropy bounds compared to existing methods and remains valid even if the preparation and measurement devices share quantum correlations. Overall, the methods developed in this work enable the integration of SDI technology into portable telecom devices, opening up a new generation of quantum cryptographic hardware.

High-stakes clinical use of large vision-language models (LVLMs) requires reasoning that is grounded in visual evidence and clinical knowledge, not just correct final answers. We introduce OpenMedReason, a large-scale, open multimodal medical reasoning corpus comprising approximately 450K image-question-answer instances whose reasoning traces are primarily derived from curated biomedical, human-authored scientific articles. OpenMedReason provides high-fidelity supervision beyond synthetic chains of thought, covering diverse medical domain vision modalities such as radiological scans, microscopic images, visible light photographs, charts, and others. We complement it with OpenMedReason-Bench, a held-out benchmark that allows fine-grained evaluation of LVLMs along three complementary axes of capability, including perception, medical knowledge, and rationale, enabling diagnostic evaluation beyond final-answer accuracy. OpenMedReason is a rich training resource that exhibits its effectiveness in both supervised fine-tuning (SFT) and reinforcement-based alignment. Training with OpenMedReason yields a 20% average improvement in VQA accuracy over the base model and achieves performance within 4.2% of the strongest comparable-scale medical LVLMs. Fine-grained performance analysis confirms that the gains are not concentrated in any single axis: OpenMedReason improves perception, medical knowledge, and rationale jointly, and its reasoning traces are preferred over those of the base model in 86.1% of pairwise comparisons. We release the code and dataset at huggingface.co/datasets/neginb/OpenMedReason.

The explosion of generative 3D assets has created a massive demand for animation, yet current motion capture methods remain brittle, restricted to species-specific templates (e.g., SMPL) or requiring labor-intensive manual rigging. We introduce TopoCap, the first unified framework capable of extracting motion from monocular video and retargeting it onto characters with arbitrary, unseen skeletal topologies, i.e., from bipeds to hexapods and inanimate objects, without test-time optimization. Our key insight is that while skeletal structures are combinatorial and discrete, the underlying physics of motion occupy a continuous, low-dimensional manifold. We materialize this insight via a two-stage generative pipeline. First, we learn a Universal Motion Manifold using a Graph CVAE that compresses heterogeneous kinematic chains into a shared, fixed-length latent code. By explicitly conditioning the decoder on a structural embedding of the target rig, we disentangle motion dynamics from skeletal topology. Second, we treat video-to-animation as a conditional flow matching problem, predicting these topology-agnostic codes from visual features. To learn this generalized prior, we introduce Mobjaverse, a massive-scale dataset curated from Objaverse-XL. Comprising over 5,000 unique skeletal topologies and 2 million frames, it exceeds the structural diversity of existing datasets by two orders of magnitude. Extensive experiments demonstrate that \MethodMotion outperforms specialist models on human and quadruped benchmarks while enabling zero-shot retargeting for the long tail of 3D creatures. Dataset is publicly available at https://huggingface.co/datasets/duckduckplz/Mobjaverse.

Unmanned aerial vehicles (UAVs) are increasingly used in inspection, search and rescue, environmental monitoring, and emergency response. However, most UAV applications still rely on pre-defined command sequences or task-specific pipelines, where developers manually connect perception, planning, flight control, simulation, logging, and safety modules. This limits the flexibility, reproducibility, and extensibility of autonomous aerial systems. This paper presents AerialClaw, an open-source software framework that enables UAVs to operate as decision-making aerial agents rather than merely command-following platforms. Given a natural-language mission, AerialClaw allows an LLM-based agent to understand the task, maintain context, invoke executable aerial skills, observe perception and runtime feedback, and iteratively update its decisions in a closed loop. The framework adopts a modular brain-skill-runtime architecture, combining hard skills for atomic UAV operations, Markdown-based soft skills for reusable task strategies, document-driven agent state and capability boundaries, memory-driven reflection, safety-oriented runtime validation, and platform-agnostic execution adapters. AerialClaw supports lightweight mock execution, PX4 SITL with Gazebo, and AirSim-based simulation, together with a web console, pluggable model backends, example missions, simulation assets, and staged deployment scripts. By combining standardized aerial skills, document-driven agent state, memory, and closed-loop LLM decision-making, AerialClaw provides a reproducible and extensible open-source framework for building UAV systems that can interpret missions, make decisions, execute skills, and adapt their behavior from feedback.

Long-video understanding remains challenging for multimodal large language models, because temporally extended videos often contain thousands of frames and are therefore expensive to process exhaustively. Existing methods usually construct compact visual inputs from long videos under a limited visual budget. However, most of them still follow a frame-centric paradigm and apply similar representations to retained content regardless of its importance. This makes it difficult to preserve both high-fidelity visual evidence and broad temporal coverage. To address this issue, we propose Q-Fold, a training-free input construction framework for long-video understanding. Instead of treating isolated frames as the basic modeling unit, Q-Fold operates on contiguous temporal segments and constructs a heterogeneous Focus--Context representation under query guidance. Query-relevant segments are preserved as high-fidelity Focus Frames, while less relevant segments are folded into chronology-preserving contextual layouts. In this way, Q-Fold preserves critical visual evidence and broad temporal coverage, while better maintaining local temporal continuity within short segments. Experiments on four long-video benchmarks with multiple Video-MLLMs show that Q-Fold consistently improves performance without increasing the input budget. Notably, it achieves gains of up to 9.1 percentage points on an ultra-long video benchmark. Code will be made publicly available.

Benchmark scores often misrepresent a large language model's (LLM's) knowledge, because they rely, e.g., on the model's ability to follow specific formatting requirements. This especially penalizes base models that may know the correct answers but lack the ability -- typically introduced in post-training -- to structure them as instructed. To overcome this, we propose soft-prompt tuning, an efficient, fair, and architecture-agnostic model evaluation. By optimizing only 10 soft-prompt vectors (roughly 0.0006% parameters for a 7B model) over a short tuning period, we adapt models to specific benchmark formats, closing gaps in format-following and ensuring that underlying knowledge is accurately reflected in benchmark scores. This allows one to fairly compare different base models -- trained with various pre-training recipes -- on benchmarks without the need for full post-training. We evaluated soft-prompt tuning across 7 models and 7 datasets. The results show that (a) soft-prompt tuning saturates format-following within 80 steps (~640 samples) making it highly efficient, (b) soft-prompt tuning significantly outperforms zero- and few-shot prompting, surfacing base model knowledge that standard prompting misses, that (c) even post-trained models can benefit from soft-prompts to maximize format compliance, and that (d) soft-prompted base model performance predicts post-trained model rankings more reliably than zero- and few-shot baselines, offering a low-cost proxy for downstream model quality. Our contributions include (1) metrics which disentangle format-following and knowledge accuracy, (2) a fairer benchmarking protocol of LLM knowledge, and (3) a cost- and memory-effective recipe to identify optimal pre-training strategies early in LLM development.

Transformer-based language models for SMILES strings suffer from a locality gap: standard character-level tokenization fragments chemically meaningful motifs, forcing models to repeatedly learn local syntax at the expense of long-range dependencies. To address this without disrupting standard tokenizers, we propose MolGram, which integrates a conditional $n$-gram memory module into molecular language models. MolGram maps local string patterns to learned embeddings via scalable hash lookups and dynamically injects this regional context into hidden states. Evaluations across three tasks, including unconditional molecule generation, forward reaction prediction, and single-step retrosynthesis, show that MolGram consistently improves performance. Crucially, our analyses demonstrate that MolGram outperforms baselines with 3$\times$ more parameters, establishing explicit local pattern memory as a highly efficient inductive bias.

Vision-Language-Action (VLA) models have demonstrated remarkable zero-shot generalization in robotic manipulation, yet the vast majority of pre-trained pipelines remain strictly confined to low-DoF parallel grippers. Adapting these rich semantic priors to high-DoF dexterous hands introduces a severe morphology gap, direct end-to-end joint fine-tuning inherently causes catastrophic forgetting of spatial reasoning and acute action manifold collapse due to data scarcity. In this paper, we present InDex, a novel, data-efficient adaptation framework rooted in cross-morphology semantic inheritance. Rather than discarding the pre-trained 1-DoF parallel grasp output, we repurpose it as a continuous, macroscopic virtual grasp intent proxy to sequentialize the control topology. We implement a two-stage decoupled learning architecture: the first stage parameter-efficiently aligns the VLA backbone to predict continuous arm trajectories and the scalar grasp intent; the second stage freezes this spatial backbone and leverages an intent-conditioned denoising diffusion head to decode fine-grained joint articulations for multi-fingered end-effectors. Extensive simulation benchmarks across a suite of multi-stage, contact-rich dexterous manipulation tasks demonstrate that InDex effectively masters intricate skills with minimal demonstration data, substantially outperforming monolithic baselines while preserving the robust spatial generalizability of the original VLA prior.

This paper presents our solution to the 2026 SoccerNet VQA Challenge. We first develop a cost-effective data synthesis pipeline driven by a Vision-Language Model (VLM), which systematically restructures raw domain data into diverse VQA samples, including concise answers and long-form responses. Second, we propose MSUE, a multi-expert question answering architecture that employs a Large Language Model (LLM) to dynamically dispatch questions to text, image, and video experts. These experts are instantiated as a strong text baseline Gemini3-Flash, a fine-tuned Qwen3-VL, and an external knowledge base, respectively, working collaboratively to enhance VQA performance. MSUE achieves an accuracy of \textbf{0.95} on the challenge benchmark, securing third place in the leaderboard.

Vision-language-action (VLA) models inherit a shared synchronous clock from vision-language pretraining, processing every input at one rate. This is misaligned with physical interaction, where a high-frequency modality changes at hundreds of hertz, vision evolves more slowly, and language stays constant across an episode. A synchronous VLA oversamples slow modalities, undersamples fast ones, and caps action generation at the lowest effective frequency. We hypothesize that decoupling temporal processing per modality, letting each update and retain information at its own sensor rate, yields stronger representations and more robust control. We present DAM-VLA, which maintains per-modality latent buffers refreshed at sensor rates and read continuously by the action head, integrating new high-frequency modalities through gated cross-attention that leaves the pretrained backbone intact. Across seven contact-rich real-world manipulation tasks, DAM-VLA more than doubles the average success rate of the strongest synchronous baseline (95.2\% vs.\ 40.95\%) while sustaining smooth, reactive 100\,Hz control. Project website: \href{https://intuitive-robots.github.io/DAM-VLA/}{intuitive-robots.github.io/DAM-VLA/}

Most fairness research in NLP assumes direct access to protected attributes such as gender, race, or nationality. In practice, however, such information is often unavailable due to privacy constraints, missing metadata, or legal restrictions, even though models may infer it from indirect textual cues. This raises a key question: can debiasing succeed without direct access to sensitive attributes? We propose H-SAL, which performs post-hoc concept and attribute erasure using self-description text as an implicit debiasing signal. To support this setting, we introduce a multi-domain Stack Exchange-based fairness benchmark for helpfulness prediction that includes both explicit and implicit signals, enabling comparison between standard debiasing with protected labels and debiasing without access to sensitive information. Across encoder and decoder-only language models, we find that implicit self-description often matches or outperforms explicit-label-based debiasing. Our results broaden representation-level fairness research and provide a new benchmark for studying debiasing under realistic data constraints.

Training deep search agents requires verifiable questions whose answers remain unavailable until sufficient evidence has been acquired through search. Existing synthesis methods often increase apparent difficulty by enriching graph structures, but structural complexity alone does not guarantee realized search difficulty: the intended search process can collapse through a cheaper identifying route. We formalize this gap with a shortcut-aware difficulty framework and identify four actionable shortcut risks: evidence co-coverage, single-clue selectivity, exposed constants, and prior-knowledge binding. To diagnose their realized effects, we use trajectory signatures including solving cost, answer hit time, and prior-shortcut rate. Guided by this framework, we introduce FORT, a Framework of Shortcut-Resistant Training-Data Synthesis. FORT constructs shortcut-resistant training data by controlling shortcut risks across entity selection, evidence graph construction, question formulation, and adversarial refinement. Experiments show that FORT induces longer pre-answer search and fewer shortcut patterns than existing open-source deep search datasets. Using the resulting trajectories, we train FORT-Searcher with supervised fine-tuning (SFT) only, and it achieves the best overall performance among comparable-size open-source search agents on challenging deep search benchmarks. Relevant resources will be made available at https://github.com/RUCAIBox/FORT-Searcher.

Time-series clustering remains challenging due to the inherent trade-off between clustering effectiveness and computational efficiency. Similarity-based methods often suffer from quadratic complexity caused by pairwise distance computations, while deep learning-based approaches typically rely on costly iterative training and a large number of trainable parameters. In this paper, we propose MSRGC-Net, an efficient time-series clustering framework that integrates multiscale reservoir computing, granular-ball-based anchoring graph construction, and consensus learning. MSRGC-Net adopts a training-free reservoir computing paradigm to extract multiscale temporal representations from raw time series without backpropagation, significantly reducing computational overhead. To capture the intrinsic structure of the resulting representations, granular-ball computing is employed to adaptively model data distributions via density-consistent regions, yielding compact and robust anchor graph representations. Furthermore, a consensus-based anchoring graph optimization strategy is introduced to effectively align multiscale reservoir representations and integrate complementary information across temporal scales. Extensive experiments on widely used univariate and multivariate benchmark datasets demonstrate that MSRGC-Net consistently outperforms state-of-the-art methods in clustering performance while maintaining superior computational efficiency.

Network Intrusion Detection Systems (NIDS) heavily utlize Machine Learning (ML) but ML models can be manipulated via adversarial attacks. These attacks add carefully crafted perturbations to network traffic data that leads to misclassifications. While prior work has demonstrated adversarial vulnerabilities in isolated settings, systematic cross-architecture as well as class and category of attack based comparisons under controlled attack conditions remain limited, leaving practitioners without clear guidance on which models to deploy in adversarial environments. This paper asks a simple question: what type of classifier architectures actually hold up when attackers try to manipulate the systems? We put three popular architectures through their paces: a 1D Convolutional Neural Network, a Long Short-Term Memory (LSTM) network, and a Random Forest (RF) ensemble. Using the ACI-IoT-2023 dataset (over 1.2 million samples spanning 12 attack types), we subject each model with FGSM and PGD adversarial attacks, which apply gradient-based perturbations in normalized feature space consistent with established adversarial ML evaluation protocols, at perturbation budgets ranging from $ε=0.01$ to $ε=0.1$. Surprisingly, Random Forest achieved near-perfect baseline accuracy (99.98\%), yet collapsed catastrophically under attack, dropping 73 percentage points at the smallest perturbation we tested. CNN, on the other hand, retained 95.5\% accuracy at $ε=0.01$ and degraded gracefully as perturbations increased. LSTM fell somewhere in between. These findings flip the conventional wisdom where high baseline accuracy means nothing if a model shatters at the first sign of adversarial pressure. For practitioners deploying intrusion detection in adversarial environments, we recommend CNN-based architectures and provide scenario-specific deployment guidance.

Pretrained video generators are promising visual world models that exhibit emergent task-solving abilities; however, their reliance on detailed textual descriptions limits their direct use for planning and decision-making. Existing approaches either outsource this reasoning to language or vision-language models, or rely on supervised fine-tuning with paired task-execution videos, which are costly to collect and difficult to scale. We propose a scalable framework that elicits task-solving ability in such models by combining self-distillation with reinforcement learning. Given an unlabeled scene image, a vision-language model generates a candidate task and a detailed step-by-step solution. The solution conditions a pretrained video diffusion model, the Demonstrator; we distill its behavior into an Executor conditioned only on the image and a short task prompt. This transfers execution knowledge from caption-guided generation to instruction-conditioned task solving without curated task-video supervision. We further improve the Executor with reinforcement learning from VLM feedback, exploiting the asymmetry between judging whether a sampled video satisfies a task and generating the solution. Experiments on our proposed WorldTasks-Benchmark and the DreamGen robotics benchmark show that the Executor surpasses the Demonstrator under our VLM-based evaluation protocol and transfers competitively to robotic tasks.

LLMs are increasingly used to generate and judge scientific ideas. This makes novelty evaluation a central problem. Full idea evaluation is difficult because it often requires judging a method, its feasibility, and its empirical promise. We therefore study a cleaner upstream object: the research question (RQ). RQ generation is a prerequisite for scientific ideation, and RQs can be compared against questions pursued in real papers. We introduce RQ-Bench, a benchmark built from recent arXiv papers. For each paper, we reconstruct author-anchored RQs from its cited background, gaps, and contributions. These RQs are not the only valid questions for the same background. They are author-anchored reference points for testing novelty judgments. We evaluate model-generated RQs with standalone LLM judging, comparative LLM judging, and human expert evaluation. LLM judges consistently rate model-generated RQs as highly novel, producing a novelty mirage; in comparative evaluations, this preference becomes even stronger. Domain experts, however, reach the opposite conclusion and prefer the author-anchored reference questions. We further find that many generated RQs are narrow or source-bound, a dimension that LLM judges often miss unless explicitly tested. Overall, the contradictory novelty evaluations between LLM judges and human experts raise a serious concern about the reliability of using LLMs to assess the scientific novelty of research questions.

Touch is the primary medium through which humans interact with the environment. Currently, tactile learning mainly focuses on image-level pretraining or alignment. However, tactile signals correspond to local object contact, while research into scale alignment and holographic matching remains limited and proper datasets and benchmarks also lack. To bridge this gap, we first construct a data collection system to acquire a large-scale tactile dataset, with over 20 K tactile contacts from 505 real-world objects. Building on this dataset, we design a Vis-Tac Holographic Matching Benchmark to evaluate vision-tactile local-to-global alignment ability. Then we propose Vision-Tactile Patch Alignment (VTPA) methods for vision-tactile representation learning. Experiments demonstrate that these exceed the performance of methods without alignment and align with whole-object images.

In modern vehicular systems, robust performance under harsh conditions has become a critical problem of autonomous driving. Our study delivers a comprehensive evaluation of the newest iteration of the YOLO series, which is YOLOv11 Nano architecture benchmarked against the widely adopted YOLOv8 Nano as a baseline on a custom fused dataset that combines the Indian Driving Dataset (IDD) [1] and Berkeley Deep Drive Dataset (BDD100K) [2]. We have analyzed the trade-offs among detection accuracy, inference speed, and computational efficiency in high-entropy scenarios involving dense mixed traffic, rain, and low-light conditions. Specifically, YOLOv11n achieves a mean Average Precision (mAP@50) of 46.6%, with a notable 3.2% improvement in Precision over the baseline, effectively reducing false positives in cluttered scenes. Furthermore, the proposed model exhibits enhanced energy efficiency, requiring 22% fewer FLOPs (6.3G vs. 8.1G) while maintaining real-time inference speed of 70.9 FPS on a Tesla T4 GPU, offering an optimal trade-off for safety-critical edge deployment.

Automating compliance check for geometry-intensive regulations remains a significant technical bottleneck in Building Information Modeling (BIM), primarily due to the semantic disparity between high-level regulatory logic and structured IFC data. Existing methods, often reliant on static rule templates, struggle to traverse multi-hop reasoning chains or resolve latent spatial dependencies across multiple building entities. To address these challenges, a Spatial-Geometric Reasoning System for Building Information Modeling (SGR-BIM) is proposed as an integrative graph-driven reasoning framework. SGR-BIM dynamically constructs a cross-modal knowledge graph that aligns user intent, regulatory semantics, and BIM geometry, enabling interpretable reasoning without rigid hard-coding. Validated on 679 expert-verified queries from fire safety codes, the framework achieves 84.3% accuracy, representing an 8.6% improvement over enhanced-tool single-agent baselines. This research provides a graph-based semantic reasoning paradigm, enhancing the transparency and flexibility of automated geometric compliance check workflows in the Architecture, Engineering, and Construction (AEC) industry.

Physics-informed neural networks (PINNs) combine machine learning with physical laws to solve differential equations. While existing results provide rigorous \emph{a posteriori} upper bounds for PINN prediction errors, complete certification also requires complementary lower information in order to obtain computable two-sided error enclosures. In this paper, we derive computable \emph{a posteriori} lower bounds for PINN errors in ordinary differential equations on suitable certified state-space domains under a localized strong monotonicity condition. We combine these estimates with complementary localized upper bounds under a one-sided Lipschitz condition, which is weaker than the global Lipschitz assumption used in previous work and can yield sharper upper error bands. The resulting bounds depend only on the neural-network approximation, the ODE residual, and local monotonicity and growth constants, and therefore do not require access to the exact solution. For linear time-invariant and time-varying systems, we further derive explicit formulas in terms of the minimal and maximal eigenvalues of the symmetric part of the system matrix. We also discuss the distinction between soft and hard enforcement of initial conditions in PINNs and explain why exact enforcement can make the scalar lower certificate uninformative. To recover nontrivial lower information in the linear setting, we use a signed-residual finite-probe certificate based on coordinate unit vectors. We also formulate a certificate-informed training strategy in which the propagated upper certificate is used as an auxiliary regularizer, while lower certificates remain post-training diagnostics. Altogether, the proposed framework provides rigorous and practically computable error certificates for PINN approximations of ODEs, while making explicit the domains and model classes for which the assumptions can be verified.

In this paper, we address the problem of zero-shot understanding of accidents from surveillance videos by identifying when an impact event occurs, what type of impact it is, and where in the frame it occurs using natural language. We propose a three-stage pipeline that decomposes the accident understanding into when, what, and where. The first stage extracts a short temporal window around the impact using vision-language similarity. In the second stage, we perform metadata-driven multi-prompt reasoning with five complementary views (baseline, motion, geometry, contrast, and tiebreaker) and resolve disagreement via an entropy-gated pairwise adjudicator. Finally, we localize the impact of an open-vocabulary detector queried on the predicted accident type and scene layout, and aggregate detections across keyframes using a score-weighted centroid. Our pipeline achieves a substantial improvement in the harmonic-mean score over a centre-of-frame baseline on the zero-shot ACCIDENT @ CVPR benchmark. We show that decomposing zero-shot video understanding into temporal localization, semantic classification, and spatial grounding enable more reliable reasoning with vision-language models than direct prompting alone.

Deep Reinforcement Learning (DRL) has shown promise for social navigation, yet its real-world deployment remains hindered by a persistent sim-to-real gap arising from simplified first-order dynamics and context-specific human state estimation pipelines. This work presents a unified framework that addresses these limitations to produce dynamically feasible navigation policies suitable for real-world deployment. First, theoretical analysis reveals that tracking error between simulated and actual robot position decays exponentially with increased control order, motivating the use of higher-order control inputs as DRL action space. A second-order control formulation tailored to differential drive robots is developed, complemented by a stochastic iterative Linear Quadratic Regulator (iLQR) that pretrains the policy via a divergence minimization objective. Second, to avoid the added system complexity of camera-LiDAR fusion, a cluster-based human tracking pipeline using only 2D LiDAR is introduced. Human detections are associated according to both spatial proximity and velocity similarity, enabling reliable differentiation of nearby pedestrians and yielding stable velocity estimates through temporal aggregation. Third, we introduce an unbiased residual gating block to balance reaction- and memory-based behaviors while handling time-varying crowd sizes, both critical for social navigation. The resulting policy, KinematicRL, consistently improves kinematic performance and adapts to varying number of detected humans. Experiments in real-world environments demonstrate that, when combined with the proposed tracking pipeline, KinematicRL can be deployed on a real differential drive robot with minimal modifications.

The design of reinforced concrete highway barriers is a safety-critical process that requires strict compliance with regulatory provisions such as the AASHTO-LRFD bridge design guidelines. Current engineering practice relies heavily on manual, iterative, and heuristic calculations to satisfy complex nonlinear material and mechanics constraints. Although Large Language Models (LLMs) demonstrate strong generative capabilities, their direct application to structural engineering remains limited by hallucination risks and insufficient physical grounding. To address these challenges, this study proposes a novel "generation-evaluation-optimization" closed-loop framework for automated concrete barrier design using the multi-agent orchestration capabilities of AutoGen. Experimental results demonstrate that the proposed agentic framework achieves over 98% design accuracy, significantly outperforming standalone general-purpose LLMs. More importantly, the study reveals that design performance is not necessarily correlated with model scale, where an 8B-parameter lightweight model could outperform unconstrained 631B-parameter flagship models. This finding highlights the potential to substantially reduce computational costs while improving the accessibility of AI-assisted engineering tools for industry applications. The source code for the proposed multi-agent design framework is available at the project GitHub repository: https://github.com/MXY820/barrier-design. Keywords: Structural Engineering; Multi-Agent Systems; Large Language Models; Concrete Barrier Design; AutoGen; Design Automation.

Recent advances in Vision Transformers (ViTs) for face recognition (FR) have moved beyond the standard CLS-token paradigm. In this paradigm, a special classification token (CLS) is prepended to the patch embeddings and used as a representation of the input for downstream tasks. An alternative approach, Concatenated Patch Embeddings (CPE), instead leverages all patch tokens by concatenating them into a single vector, which is then projected into a compact face representation. CPE has been shown to improve recognition performance in comparison to CLS-based ones, but our qualitative analysis of attention maps showed the presence of artifacts that limit their interpretability. To address this issue, we incorporate register tokens, learnable tokens concatenated to the initial patch embeddings, and processed jointly through the ViT encoder blocks. This mechanism has been shown to produce more structured and interpretable attention maps compared to baseline ViT. We empirically demonstrate that these artifacts consistently appear across various ViT backbones, including small and large models, and that introducing register tokens effectively mitigates them. Adding four or eight registers significantly enhances interpretability, with eight registers providing the highest verification accuracies and smoothest attention structures. Our resulting model, ViT-8R, corresponds to a CPE-based ViT-B architecture augmented with eight register tokens achieves state-of-the-art performance among ViT-based FR models on large-scale IJB-B and IJB-C benchmarks. Also, ViT-8R produces substantially clearer attention maps compared with the baseline model, which offer deeper insight into the model's attention behavior (https://github.com/TaharChettaoui/ViT-FR-Registers)

Generalizable robot manipulation requires not only task-level reasoning over unseen scenes, but also reliable grounding of visual plans into embodiment-specific execution. To bridge this gap, we propose VICX (Video generation and In-Context eXecution), a decoupled closed-loop manipulation framework. In VICX, a frozen video generation model produces vision-language-conditioned high-level visual plans, while a Video-to-Trajectory In-Context Operator Network (V2T-ICON) serves as the task-agnostic interface that grounds these plans into executable robot-state trajectories. To improve execution generalization, V2T-ICON operates on segmentation-extracted arm-only frame observations and uses retrieved image-state pairs as in-context prompts, allowing a robust and generalizable visual-to-state mapping at inference time without parameter updates. Experiments on Meta-World show that VICX supports cross-task generalization, closed-loop self-correction, and cross-embodiment transfer, demonstrating dual generalization across both task semantics and robot execution. The project webpage can be found here: https://scaling-group.github.io/vicx/.

Collision-free path planning in cluttered, real-world environments relies on a representation of the collision-free space, and existing representations broadly fall into two categories. Explicit representations, such as unions of convex sets, can be plugged into optimization-based planners as hard collision-free constraints, but their parameters scale poorly with configuration-space dimension. Implicit representations, by contrast, are flexible and scale well to complex geometries, yet typically lack such guarantees. We bridge this gap with ILD (Invertible Latent Decomposition), a framework that jointly learns an invertible mapping and a union of explicit convex polytopes in the resulting latent space. Planning is carried out over these latent convex sets, and the invertible mapping decodes the resulting paths back to the original configuration space while preserving feasibility with respect to the refined explicit safe regions. We further propose Visibility-Guided Sampling (VGS) to keep the convex sets connected for path planning. Across 2D navigation, 6-DoF, and 14-DoF manipulation environments, ILD achieves broader coverage, better inter-set connectivity, and higher path-planning success rates than prior baselines, with zero observed false positives after test-time refinement. On a 14-DoF bimanual manipulator, we further demonstrate real-time collision-free planning, with test-time refinement adapting to scene-geometry changes during real-world deployment on a single 6-DoF arm.

Finite element (FE) modeling of safety-critical infrastructure such as bridge barriers requires high-fidelity nonlinear dynamic analysis, yet the current FE modeling process remains labor-intensive and lacks automation. This paper presents the Human-Enhanced Loop Modeling (HELM) framework, a collaborative human-agent protocol that decomposes long-sequence finite element modeling into discrete, visually verifiable checkpoints across geometry generation, boundary condition definition, and material assignment. The framework is demonstrated through a 20-case matrix of reinforced concrete bridge barriers under MASH TL-4 and TL-5 lateral loading conditions, interfacing specialized agents with two widely used commercial FE softwares, i.e., ANSYS and LS-PrePost. Experimental results show that HELM improves the baseline autonomous modeling success rate from 20% to 75%, with agent-level pass rates for geometry and boundary condition tasks approximately doubling. Error analysis reveals that spatial reasoning and algebraic logic limitations constitute the primary failure modes, underscoring the value of structured human-in-the-loop intervention for modeling automation. The complete agent design code and prompts are open-sourced and can be accessed at: https://github.com/SimAgentDev/Ansys-LSPP-AgentKit.

Vision Transformers (ViTs) have gained significant attention in computer vision and shown strong potential for face recognition (FR). However, their high computational cost makes deployment on resource-constrained devices challenging, motivating the need for methods that balance efficiency and accuracy. In this work, we investigate early exiting in pretrained ViTs as a simple yet effective training-free strategy for efficient FR inference. Leveraging the uniform feature dimensionality across transformer encoder blocks, we introduce ViT-FREE, a multi-exit framework that enables face verification directly from intermediate representations without modifying or retraining the backbone model, and thus, reducing inference cost. Empirically, we show that patch embeddings and attention maps evolve progressively across depth, exhibiting high similarity between consecutive ViT blocks and increasing alignment with the final representation. This indicates gradual feature refinement and attention convergence, suggesting that intermediate layers already provide stable and discriminative representations suitable for early exiting. Through extensive experiments on multiple FR benchmarks, we systematically analyze the accuracy-efficiency trade-off across exit depths. Our results demonstrate that later exits achieve a highly favorable balance, with exiting at layer 10 yielding up to a 20% speedup while incurring only a 1.5 drop in verification performance on benchmarks such as IJB-C. Also, we propose ViT-FREE_FT, a lightweight exit-specific fine-tuning strategy that adapts only the projection layers using a small synthetic dataset while keeping the transformer backbone frozen. This approach improves the performance of shallow exits while preserving the efficiency benefits and leaving deeper exits largely unaffected.

We propose a multi-agent collaborative framework built upon a lightweight Multimodal Large Language Model (MLLM), specifically designed for social intelligence reasoning. A key feature of our approach is that both the training and inference phases are augmented via knowledge distillation. Within this architecture, multi-modal data pertinent to social intelligence is precisely localized. Furthermore, relevant long-tail events are identified, extracted, and rendered as formatted, explicit text. This formatting strategy prevents critical long-tail information from being overshadowed by head events and environmental noise during the tokenization process. Specifically, we integrate Test-Time Adaptation (TTA) across the entire reasoning pipeline, encompassing the extraction and representation of long-tail events, Chain-of-Thought (CoT) prompting, and self-reflection. This TTA mechanism is also distillation-enhanced, utilizing Low-Rank Adaptation (LoRA) to fine-tune the foundation model exclusively for instance-level reasoning. Extensive evaluations against various open-source and proprietary AI models across multiple benchmarks demonstrate the effectiveness of the proposed framework. With around 30% of training data from IntentTrain, we achieve state-of-the-art results. Codes are available at https://github.com/eeee-sys/MODF-SIR, demo is available at https://huggingface.co/spaces/Harry-1234/MODF-SIR, LoRA is available at https://huggingface.co/Harry-1234/MODF-SIR and the dataset for training router is available at https://huggingface.co/datasets/Harry-1234/IntentRouterTrain.

Model post-training, and in particular reinforcement learning (RL), is one of the primary mechanisms by which developers can shape models' values and behaviors. However, as models become increasingly evaluation and training aware, they may be motivated to resist training when the perceived objective conflicts with their current values, undermining developers' ability to detect misalignment and correct model behavior through further training. In this paper, we demonstrate generalization hacking, in which a model collects reward during RL while preventing the rewarded behavior from generalizing. We construct a model organism on Qwen3-235B-A22B, finetuning on synthetic documents describing training awareness and self-inoculation, a novel mechanism in which the model frames compliance as context-specific in its chain of thought, without demonstrating or instructing either behavior. The model organism achieves train-time harmfulness comparable to controls while maintaining a persistent ${\sim}15$ percentage point compliance gap across 700 steps of RL. Additionally, a control organism trained only on training awareness documents independently discovers inoculation-like reasoning under RL pressure, developing its own compliance gap despite never being exposed to the concept. Because the generalization-hacking organism receives high reward throughout, standard training metrics provide no signal that generalization has failed. Our results constitute the first demonstration that a model can actively resist RL behavioral modification while maintaining high reward, suggesting that as models become more capable and training-aware, they may be able to undermine the training process itself.

Predicting time-to-event outcomes such as mortality is a fundamental task in clinical decision-making, commonly addressed through survival analysis. While classical statistical and deep learning approaches have been widely studied, they typically require task-specific training and sufficient labeled data. Recent advances in tabular foundation models offer a new paradigm by learning general-purpose representations for structured data. However, their applicability to censored time-to-event prediction in clinical settings remains underexplored, as typical applications are restricted to discrete classification rather than survival analysis tasks. In this work, we propose a lightweight adaptation approach for applying tabular foundation models to clinical survival analysis by directly training a survival-aware head on top of the pretrained representations. We study representative architectures, including TabPFN, TabDPT, and TabICL, and adapt them using a multi-task logistic regression (MTLR) head to model right-censored time-to-event outcomes. We evaluate this approach on a diverse set of public survival benchmarks and two large-scale ICU cohorts, MIMIC-IV and eICU. Our results show that this transfer learning approach achieves competitive or superior performance compared to strong baselines. On MIMIC-IV, TabDPT-FT-MTLR reaches a C-index of 0.856, corresponding to a relative improvement of +1.4% over the best non-FM baseline (DeepSurv, 0.844) and +6.7% over the best zero-shot model (0.802). On eICU, TabICL-FT-MTLR achieves 0.797, yielding gains of +1.7% (DeepSurv, 0.784) and +6.4% (0.749), respectively. These findings highlight the importance of combining pretrained tabular representations with survival-aware objectives and suggest that tabular foundation models provide a practical and effective alternative for clinical survival prediction.

Trusted monitoring is a cornerstone of AI control. However, as frontier models grow more capable, the increasing capabilities gap between trusted and untrusted models may render trusted models unreliable monitors. We introduce \emph{bootstrapped monitoring}, a protocol that addresses this by inserting a stronger, intermediate untrusted model with transparent chain-of-thought reasoning into the oversight chain. The untrusted monitor ($U_m$) evaluates the agent's actions, while a weaker trusted model ($T$) oversees $U_m$'s reasoning to detect collusion. We evaluate bootstrapped monitoring on multi-turn software engineering tasks (BashArena) across multiple agents and monitors. Bootstrapped monitoring substantially improves catch rates over trusted-only monitoring, even when the untrusted monitor actively colludes with the agent, provided we have access to its raw chain-of-thought. Our results suggest that bootstrapped monitoring can extend the useful lifetime of trusted models in control as AI capabilities advance.

Remaining Useful Life (RUL) prediction is essential for industrial predictive maintenance, yet many learning-based approaches rely on extensive feature engineering or large labeled datasets to train task-specific sequence models. In this work, we introduce a lightweight learning approach, in which we leverage a frozen pretrained time-series foundation model (TSFM) and combine it with a small regression head for RUL estimation from multivariate sensor streams. More specifically, we use Chronos-2 as a frozen backbone to extract context window features and train a lightweight regression neural network for RUL prediction. Experiments on real-world industrial sensor data from two device types show that Chronos-2 features consistently improve over recurrent, convolutional, Transformer-based, and gradient-boosting baselines under the same preprocessing and evaluation protocol. We further analyze the impact of context length and find that performance improves significantly with longer histories, indicating that TSFM representation offer a practical and data-efficient alternative for RUL estimation in industrial settings.

Credible simulated-rainfall conditions are essential for identifying perception-system boundaries and supporting SOTIF-oriented risk assessment in automated driving. However, closed-field tests are often described only by nominal rainfall intensity or single-point measurements, making it difficult to align simulated rain fields with real rainfall and map test results to real-world scenarios. This paper proposes a path-based credibility evaluation method for simulated rainfall in autonomous-driving perception tests. Using the drop size and velocity joint distribution of real rainfall as the reference, each candidate path is represented by path-equivalent rainfall intensity, an uncertainty band, and a path-averaged Realism of Raindrop Distribution (RRD) score. Lidar target point-cloud count and mean reflectivity are further used for perception-consistency correction, quantifying the proxy capability of each simulated-rainfall path for real-rainfall perception effects. Experiments are conducted using about 10,000 real-rainfall raindrop-spectrum samples, 728 RainSense perception samples, and 45 spatial sampling points in a 2.4 m x 7.2 m simulated-rainfall area. Results show that spatial non-uniformity remains under the same nominal condition, confirming the need for path-based evaluation. The method identifies Path IV and Path VI as preferable candidates, with results of 11.54 +/- 0.31 mm/h, RRD = 0.43, and 8.28 +/- 0.34 mm/h, RRD = 0.46, respectively. These paths show more balanced performance in rainfall-intensity stability, raindrop-spectrum realism, and perception consistency. The proposed method supports path selection, condition description, and credible interpretation of autonomous-driving perception tests under rainfall.

Preference-based reinforcement learning (PbRL) learns policies from human trajectory-level comparisons, avoiding explicit reward design and expert demonstrations. Existing methods typically train utility functions on trajectory or segment-level preferences while relying on per-step utility estimates during policy optimization. This training and inference mismatch induces a distribution shift that severely degrades temporal credit assignment and limits policy learning. We analyze this issue and propose PAWS, a segment-based preference learning method that performs policy updates directly using segment-level advantage functions. By aligning utility training with policy optimization, PAWS preserves trajectory-level preference information and avoids unreliable per-step learning signals. Experiments on simulated robotic manipulation and locomotion tasks demonstrate that PAWS consistently outperforms existing PbRL approaches, highlighting the importance of distribution-consistent preference learning.

In this report, we present our third-place solution for the DataMFM Challenge Track 1: Document Parsing. This track requires models to recover structured Markdown documents from document page images while preserving textual content and document structure. To address the complementary requirements of accurate content recovery and faithful structure reconstruction, we propose ParseFixer, an agentic framework for backbone parsing and selective correction. ParseFixer consists of two key modules: Full-Page Backbone Parsing (FBP) and Agentic Selective Correction (ASC). FBP produces stable initial Markdown outputs with MinerU2.5 Pro, while ASC detects high-value parsing failures and repairs them through a verify-and-rollback correction process. By placing selective multimodal correction after open-source backbone parsing, ParseFixer improves the recovery of key document elements without rewriting reliable backbone predictions. On the test set, our final system achieves an overall score of 61.78 and ranks third in Track 1, demonstrating its effectiveness for accurate document parsing. Our code will be released at: https://github.com/iLearn-Lab/CVPRW26-ParseFixer.

Software engineering tools increasingly rely on LLM based agents to localize files to change to resolve a software issue. Most AI agents explore repositories linearly, that is, visiting one directory or file per step. We postulate that this is a structural mismatch for changes that span several subsystems. We compare linear sequential exploration against non-linear, domain-scoped parallel agentic exploration. Using SWE Bench Pro as initial benchmark, we focus on ansible as an exemplar. We construct an approach for persistent-session evaluation of GitHub issues anchored at a single base commit. We compare our non-linear domain-agent file traversal system against a base LLM without direct repository access, a single agent Recursive Language Model (RLM) baseline with a persistent Python REPL and an external CLI baseline using Codex 5.5 High. Domain scoped parallel agent spawning with a small Haiku-class model achieves the highest micro F1 among Haiku class models by a large margin. Domain-agents is the second highest behind only the much larger Codex 5.5 High on our own expanded benchmark including over more recent PRs from 2025 and 2026. On the original, curated, 2020 SWE-bench Pro benchmark, a larger Sonnet plain LLM baseline attains higher micro F1 by predicting few files, leading to higher precision, but at significantly lower all gold recall. We also present three additional findings. First, documentation evolution is a latent dependency unresolved by any approach. Second, naive file system access can degrade localization driven by test-file over prediction. Lastly, forced multi-agent consultation does not measurably help and raises token cost substantially.

Flow Matching has enabled robust text-to-video generation via latent ODE sampling. However, velocity approximation and numerical discretization errors inevitably accumulate, causing sampling trajectories to drift. Consequently, generated videos often suffer from severe spatiotemporal inconsistencies. Nevertheless, directly correcting these drifted, noisy latents is challenging: (i) timestep-dependent noise obscures reliable structural cues; (ii) spatial interventions risk disrupting intricate local geometry while incurring heavy computational costs. To address this, we propose Spectral Lookahead Rectification (SpecLoR), a plug-and-play inference method that bypasses noise via lookahead prediction, and circumvents spatiotemporal entanglement by shifting corrections to the frequency domain, where universal statistical priors of natural videos are readily available. First, during early sampling stages, SpecLoR looks ahead to estimate the clean latent $z_{t,0}$ and computes its 3D spatiotemporal spectrum. Next, SpecLoR rectifies the amplitude spectrum to match the prior, leaving the phase intact. Finally, the corrected state is re-noised to resume ODE integration. Experiments on Wan2.2 demonstrate that SpecLoR significantly reduces physical artifacts and enhances motion coherence across multiple benchmarks with minimal computational overhead (4 additional NFEs).

This paper introduces a novel tactile sensor for in-hand manipulation with slip-aware control that integrates velocity, force/torque, and pressure map sensing into a single device with a deformable contact pad. To the best of our knowledge, this is the first sensor to combine these sensing modalities within a single compliant structure. The sensor features a deformable contact surface and can robustly track both flat and curved surfaces across a wide range of materials. Its performance is evaluated through a comprehensive set of experiments that highlight both its capabilities and limitations. The sensor is designed for rapid and low-cost fabrication using a combination of standard PCB manufacturing and rapid prototyping techniques.

This report describes our participation in SemEval-2026 Task 8 on multi-turn retrieval and question answering. The task evaluates conversational systems across four domains (finance, cloud documentation, government, Wikipedia), and includes unanswerable queries where the available collection does not contain sufficient evidence to produce a complete response. We propose a multi-turn retrieval-augmented generation pipeline that combines learned sparse retrieval with LLM-based reranking and generation. Using sparse retrieval as the primary retrieval method, we leverage its strong generalization across domains. In addition, we make use of the long-context capabilities of LLMs for conversational query rewriting, pointwise and listwise reranking, and generating the final response, each conditioned on the full conversational history. This multi-step design enables effective integration of conversational context throughout retrieval and generation, improving robustness across domains.

The native processing of time-dependent signals from optical sensors by integrated photonic circuits can potentially bring significant advantages in terms of energy consumption, latency and processing power, as it allows skipping or reducing the use of fast digital electronics and directly exploiting optical degrees of freedom and parallelism. However, due to a short memory, optical operations usually struggle to directly process optical signals with relatively slow (<MHz) dynamics from optical sensors. In this work, we experimentally show that these limitations can be overcome by exploiting the self-pulsing dynamics in a microring resonator (MRR) network. In particular, we demonstrate that such dynamics can expand and retain information about perturbations sensed by a fiber sensor. This reduces the minimum sampling rate for the digitization of the sensor signal by at least one order of magnitude. The reduction is achieved by combining fiber sensing measurements at two different perturbation locations and frequencies with MRR network measurements at multiple output ports, input power levels and laser wavelengths. This work represents a first step in bridging time-dependent optical processing and optical sensing at sub-μs time scales.

Scientific progress depends on a repeated loop of exploration, experimentation, and abstraction. Researchers test candidate directions, interpret the evidence, and carry the resulting lessons into later attempts. We study how an AI agent can run this loop autonomously over long horizons. We introduce Arbor, a general framework for autonomous research that combines a long-lived coordinator, short-lived executors, and Hypothesis Tree Refinement (HTR), a persistent tree that links hypotheses, artifacts, evidence, and distilled insights across time. The coordinator manages global research strategy over the tree, while executors implement and test individual hypotheses in isolated worktrees. As results return, Arbor updates the tree, propagates reusable lessons, refines the search frontier, and admits verified improvements. This design turns autonomous research from a sequence of local attempts into a cumulative process in which strategy, execution, and evidence are carried across time. We evaluate Arbor under Autonomous Optimization (AO), an operational setting where an agent improves an initial research artifact through iterative experimentation without step-level human supervision. Across six real research tasks in model training, harness engineering, and data synthesis, Arbor achieves the best held-out result on all six tasks, attaining more than 2.5x the average relative held-out gain of Codex and Claude Code under the same task interface and resource budget. On MLE-Bench Lite, Arbor reaches 86.36% Any Medal with GPT-5.5, the strongest result in our comparison.

Recent respiratory sound classification (RSC) studies largely rely on CLS-token driven self-attention architectures such as the Audio Spectrogram Transformer (AST). While effective at modeling global context, recent analyses suggest a low-pass filtering behavior that may reduce sensitivity to localized abnormal patterns. In this work, we investigate State Space Models (SSMs) as an alternative backbone for RSC. Using the Distilled Audio State Space model, we analyze intermediate representations through spectral response curves and observe stronger preservation of mid-to-high spatial-frequency components. Based on these observations, we introduce spectral-aware layer regularization using Gaussian convolution applied to selected layers. We further propose Dual-Axis Patch-Mix contrastive learning tailored to SSM-based audio models for robust representation learning. Experiments on the ICBHI benchmark show that our approach achieves 64.48% score, outperforming the AST baseline by 5%. Code is available at https://github.com/RSC-Toolkit/Lung-SRAD.

This paper proposes an empirical methodology to study software aging in GPU-based LLM serving systems. Traditional aging studies focus on CPU-centric software with relatively regular workloads; LLM serving is different, spanning a Python host and a CUDA device, handling requests whose cost varies by orders of magnitude, and relying on rapidly evolving software stacks. We run a 216-hour campaign across six co-located deployments under identical stress conditions, monitor host, device, and client metrics in parallel, and apply a statistical pipeline that accounts for autocorrelation and multiple testing. Our results reveal statistically significant memory aging in all deployments, with leak rates strongly dependent on the serving runtime and deployment configuration. Beyond these findings, we provide a reproducible framework that opens a research direction at the intersection of the software aging and rejuvenation and LLM serving communities.

CSI-based localization with spatially distributed antenna arrays exposes a basic resource trade-off. Each array can provide a rich view of the channel, but forwarding observations from all arrays to a fusion center is wasteful when only a few carry useful information, and the shared uplink supports only a limited number of simultaneous transmissions. We let each array decide locally whether its current observation is worth reporting, subject to a budget on the average number of active transmitters. We refer to this abstraction as Edge-Triggered Distributed Inference (ETDI). It captures a broader class of task-oriented communication problems where resource-constrained devices share an access channel for a common inference task. We instantiate ETDI for CSI-based localization, a common scenario in vehicular IoT networks. Spatially distributed remote antenna arrays (RAAs) encode local channel state information (CSI) from user equipment (UE) transmissions into latent features, and the fusion center estimates the UE position from the subset of reported features. We propose NARRAS, a decentralized reporting policy in which each RAA combines a recurrent summary of its recent observations with a memory of the last latent it transmitted. Training controls an explicit activity budget through differentiable activity penalties and validation-calibrated deterministic thresholds, and uses channel-chart regularization to shape the latent geometry. Experiments show that, at comparable uplink activity, NARRAS improves localization accuracy over learned and heuristic sparse-reporting strategies, while dense full-report models remain useful budget-free references. In low-activity regimes, chart regularization further reduces high-percentile localization errors, suggesting that geometry-aware latent representations are more robust under sparse reporting.

We propose a new paradigm for long video understanding by treating a long video as a Neural Knowledge Representation (NKR). NKR represents video contents neither as a stream of tokens nor pre-organized databases, but as an individual small portion of network weights attached to the VLM backbone. The NKR weights are optimized to encapsulate the video's semantic content via a novel Agentic Knowledge Distillation (AKD) process, where an agent automatically synthesizes dense descriptions and question-answer pairs to distill the video's knowledge into the NKR. While AKD serves as a comprehensive, one-time encoding phase, the resulting NKR transforms the video into a portable, reusable asset. At inference, the lightweight NKR is mounted onto a frozen Vision-Language Model (VLM), enabling direct, query-based understanding without reloading or re-encoding the original video. This approach decouples video length from inference cost, offering high amortized efficiency for multi-turn video understanding. Experiments on the LVBench benchmark show our method achieves performance comparable to state-of-the-art approaches while reducing end-to-end latency by over two orders of magnitude, opening new possibilities for interactive long-video understanding.

Persistence diagrams are common representations in topological data analysis, but they do not naturally live in a vector space, and the statistical tools developed for comparing them have largely evolved separately from those used for downstream prediction. We introduce STRAND (Survival Topological Representation ANalysis of Diagrams), which treats (collections of) PDs as survival data: each topological feature with persistence value $p = d - b$ is a fully observed time-to-event, and the persistence survival function $S(t) = \mathbb{P}(p > t)$ is the central object for comparing diagrams. From this single representation we derive (i) a non-parametric two-sample test with calibrated Type I error and high power from a small number of diagrams; (ii) interpretable effect sizes; and (iii) a 1-Wasserstein-stable feature vector for downstream machine learning. We validate calibration and power on synthetic manifolds with controlled topology, demonstrate competitive vectorisation across 14 graph and 3D point cloud benchmarks, and apply the method to study functional brain connectivity in fMRI/neuroscience data. To our knowledge, STRAND is the first method to provide hypothesis testing and vectorisation for persistence diagrams from a single coherent and interpretable representation.

Traffic law liability determination is critical for assigning legal penalties, requiring the simultaneous identification of interdependent statutory provisions across multiple legal dimensions. However, existing retrieval-augmented generation methods suffer from a multi-dimensional retrieval bottleneck: single axis architectures compress complex legal queries into a single pathway, causing interdependent statutory dimensions to be overlooked. To address this, we propose OMAGR, an ontology-guided framework that decomposes queries into ontology-aligned anchors and executes parallel graph retrieval across each dimension, ensuring independent retrieval across dimensions before fusion. To evaluate the proposed method, we created the TrafficLaw-QA dataset, an expert-validated benchmark dataset containing 200 questions and 527 legal provisions. Results show that TrafficOmni-RAG outperforms baselines on Context Precision and Faithfulness metrics. The findings demonstrate that parallel multi-anchor retrieval effectively resolves the multi-dimensional retrieval bottleneck, offering a promising direction for traffic law liability determination research.

Benchmarks are essential for evaluating embodied spatial intelligence, yet their construction is labor-intensive, hard to reuse, and difficult to maintain. Existing embodied benchmarks are often static and may quickly become saturated as models improve, limiting their ability to distinguish new capabilities. We propose Embodied-BenchClaw, an autonomous agentic system for constructing embodied spatial intelligence benchmarks. Given a user-specified evaluation intent, Embodied-BenchClaw automatically produces a complete and continually updatable benchmark package through a five-stage pipeline: intent blueprinting, data collection, structuring and cleaning, benchmark synthesis, and evaluation reporting. The pipeline is coordinated by three agents for planning, construction, and evaluation. To improve reusability and reliability, Embodied-BenchClaw introduces an extensible Skill Library and process quality control, enabling benchmark construction to be composable, verifiable, and repairable. We instantiate multiple benchmarks covering indoor spatial reasoning, outdoor spatial reasoning, robotic manipulation, quadruped robot navigation, UAV/aerial-view understanding, and static benchmark enhancement. These benchmarks span diverse embodied carriers, data sources, and spatial capabilities. Experiments with human evaluation, judge-based assessment, consistency checks, cost analysis, and ablations show that Embodied-BenchClaw can construct verifiable, executable, maintainable, and diagnostically useful embodied spatial benchmarks with reduced manual effort.

Vision-Language-Action (VLA) models have shown strong performance in language-conditioned robotic manipulation, yet their robustness to linguistic variation remains poorly understood. In this work, we present the first systematic multilingual evaluation of VLA models by translating the LIBERO benchmark into ten languages, revealing severe performance degradation under non-English instructions, with success rates dropping by 30-50%. Through fine-grained analysis of task executions, we find that language influence is highly non-uniform across steps: certain steps exhibit strong language dependence and dominate overall task failure, while others are largely language-agnostic. Based on this insight, we propose a step-wise inference-time intervention that aligns representations according to step language sensitivity, substantially improving performance under linguistic variation. Our results indicate that language robustness in VLA models is fundamentally a step-wise control problem, highlighting the importance of temporally structured analysis for reliable embodied agents.

Bimanual robot systems substantially expand manipulation capabilities, but coordinating two arms introduces additional control complexity and failure modes that are not well captured by existing benchmarks. We introduce DuoBench, an extensible benchmarking framework for bimanual manipulation policies on the FR3 Duo platform. DuoBench comprises eleven tasks spanning four coordination categories, implemented in simulation and partially reproduced in the real world through reproducible task recipes with 3D-printable assets. In addition, we propose a stage-based evaluation scheme that supports fine-grained semantic failure analysis beyond binary success and provide human-teleoperated datasets for all benchmark tasks. We benchmark several dual-arm imitation-learning and vision-language-action policies in simulation and on real hardware. Our results show that current policies remain challenged by bimanual manipulation, particularly in early interaction stages, parallel arm execution, and transfer between simulation and real-world settings. DuoBench provides a reproducible testbed for diagnosing these failure modes and studying future methods for dual-arm policy learning. Code, datasets, and videos are available at https://duobench.github.io/

Research on Text-Attributed Graphs (TAGs) has gained significant attention recently due to its broad applications across various real-world data scenarios, such as citation networks, e-commerce platforms, social media, and web pages. Inspired by the remarkable semantic understanding ability of Large Language Models (LLMs), there have been numerous attempts to integrate LLMs into TAGs. However, existing methods still struggle to generalize across diverse graphs and tasks, and their ability to capture transferable graph structural patterns remains limited. To address this, we introduce the GraspLLM, a framework that combines Graph structural comprehension with semantic understanding prowess of LLMs to enhance the cross-dataset and cross-task generalizability. Specifically, we represent node texts from different graphs in a unified semantic space with a frozen general embedding model, on top of which we perform motif-aware contrastive learning across multiple motif-induced adjacency matrices to extract dataset-agnostic structural information. Then, with our proposed optimal contextual subgraph, we extract the most contextually relevant subgraph for each target node and align these subgraphs to the token space of LLM via an alignment projector. Extensive experiments on TAG benchmark datasets spanning diverse domains reveal that GraspLLM consistently outperforms previous LLM-based methods for TAGs, especially in zero-shot scenarios, highlighting its strong generalizability across different datasets and tasks. Our code is available at https://github.com/Heinz217/GraspLLM.

Scientific discovery workflows usually contain and rely heavily on lab notes, where researchers record observations, interpret uncertain results, and plan follow-up experiments. Such informative lab notes preserve evolving scientific reasoning and author uncertainty, rather than polished final results exhibited in publications, providing a valuable opportunity for AI to engage in scientific exploration at a more comprehensive and deeper level. However, most prior work on scientific text focuses on papers, protocols, or structured databases, leaving informal laboratory notes underexplored as inputs to AI agents for science. This gap matters because lab notes often intermingle validated observations, tentative judgments, and possible experimental next steps within the same passage. If these signals are conflated, an AI agent may mistake uncertain scientific judgments for confirmed conclusions or executable actions. To this end, we present Notes2Skills, a two-stage framework for turning lab notebooks into verifiable skills for scientific AI agents while preserving the author's certainty. Across seven conditions and three wet-lab sessions, Notes2Skills is the only configuration that neither mistakes uncertain notes for firm instructions nor discards firm ones. We show that certainty preservation is the missing piece between lab notebooks and reliable agent skills, opening a path toward safer AI co-scientist systems.

Multi-objective reinforcement learning for humanoid robots must coordinate locomotion and manipulation within a single policy. A natural design choice is whether to use a single (unified) critic that estimates the combined value of all objectives, or separate (dual) critics with disjoint reward signals. We present a controlled comparison on the Unitree G1 humanoid (23 active DoF) in NVIDIA Isaac Lab, training loco-manipulation policies through a sequential curriculum spanning 13 levels from stationary reaching to walking with variable-orientation targets. In standardized evaluation, dual-critic policies reach targets 3.5$\times$ faster (6.5 vs. 22.6 simulation steps), achieve 2$\times$ higher throughput (14.3 vs. 7.0 validated reaches per 1,000 steps), and attain higher validated reach rates (65.2% vs. 53.8%) compared to the unified-critic policy. Notably, additional anti-gaming reward mechanisms provide no further improvement beyond the architectural change alone (60.9% vs. 65.2%). These results have direct implications for the emerging paradigm of RL fine-tuning of imitation-learned policies: when refining a pre-trained manipulation policy with RL, a unified critic risks suppressing the learned behavior through competing locomotion gradients. These findings demonstrate that critic architecture is a primary - and often overlooked - design choice in multi-objective humanoid RL, with greater impact than reward engineering on reaching efficiency.

Vision-language models (VLMs) are increasingly used for scene understanding in autonomous driving, but robustness analysis often relies on task-agnostic embedding stability alone. We study whether corruption-induced embedding drift predicts changes in a task-aligned hazard score derived from CLIP image-text similarities. Using controlled corruptions on BDD100K road scenes, we compare embedding drift against margin drift, defined as the change in hazard score under perturbation. The relationship is highly corruption-dependent: some families exhibit strong coupling between representation drift and decision drift, while others induce hazardous decision instability despite relatively modest embedding change. Furthermore, corruption families differ in failure direction: most suppress hazard detections via false negatives, while occlusion instead triggers false alarms, suggesting that benchmark design should account for asymmetric failure modes, not just overall instability rates. These results suggest that robustness benchmarks should include task-aligned stability measures in addition to embedding-level perturbation statistics.

Visual localization in complex indoor environments remains a critical challenge for robotics and AR applications. Sequential localization, where pose estimates are refined over time, is important for autonomous agents. However, traditional methods often require storing extensive image databases or point clouds, leading to significant overhead. This paper introduces a novel, lightweight approach to sequential visual localization using 3D scene graphs. Our method represents the environment with a compact scene graph, where nodes represent objects (with coarse meshes) and edges encode spatial relationships. For each image in the localization phase, we extract per-patch semantic features, predicting object identities. Localization is performed within a particle filter framework. Each particle, representing a camera pose, projects the coarse object meshes from the scene graph into the image, assigning object identities to patches based on visibility. The similarity of the per-patch features, in the input image, and object features from the scene graph determines the weight of a particle. Subsequent images are incorporated sequentially, refining the pose estimate. By leveraging a compact scene graph and efficient semantic matching, our method significantly reduces storage while maintaining performance on real-world datasets. The code will be available at https://github.com/DmblnNicole/sg2loc.

Emotional validation - explicitly acknowledging that a user's feelings make sense - has proven therapeutic value but has received little computational attention. Emotional validation in dialogue systems can be decomposed into (i) validating response identification, (ii) validation timing detection, and (iii) validating response generation. To support research on all three subtasks, we release M-EDESConv, a 120k English-Japanese multilingual corpus created through hybrid manual and automatic annotation, and M-TESC, a multilingual spoken-dialogue test set. For timing detection, we propose MEGUMI, a Multilingual Emotion-aware Gated Unit for Mutual Integration, that fuses frozen XLM-RoBERTa semantics with language-specific emotion encoders via cross-modal attention and gated fusion. MEGUMI shows superior performance on both the M-EDESConv and M-TESC datasets, both objectively and subjectively. Finally, our EmoValidBench benchmarks of GPT-4.1 Nano and Llama-3.1 8B indicate that current LLMs generate contextually similar and diverse validating responses, but emotional understanding remains a major area for improvement. Project page: https://github.com/zihaurpang/Multilingual-Emotional-Validation

Machine learning is increasingly applied to accelerate the discovery of novel materials by exploring large compositional and structural design spaces. Yet, the scarcity of high-quality data and the frequent need for out-of-distribution prediction introduce substantial uncertainty, making the assessment of model reliability essential. In this work, we investigate uncertainty quantification as a means to evaluate model confidence in the context of permanent magnet research. In a first study, we benchmark classical and modern machine learning models for predicting intrinsic magnetic properties, focusing on the quality of their uncertainty estimates. We apply Gaussian negative log-likelihood loss and dropout-based Bayesian approximation as practical strategies for estimating predictive uncertainty. In a second study, we transfer these architectural features for uncertainty estimation to a more complex task: predicting coercivity from microstructural information using a graph neural network. Together, these studies demonstrate that uncertainty quantification not only enhances the trustworthiness of predictions but is also transferable across different modeling tasks.

Custom AI agents areagents that live inside their own application, talk to their own data and tools, enforce their own security boundaries, and carry their own brand and audit trail. What separates them from the general-purpose tier is fit, not capability: each is built for one job, by the engineer who will maintain it. No published practice sets out how to build one end to end. The pieces are everywhere (function-calling APIs, the Model Context Protocol, code agents to pair with), but the practice that chains them lives in podcasts, blogs, and leaked system prompts. This paper writes that practice down as a methodology, Agents All the Way Down: two preconditions crossed once and kept, then three practices repeated for the agent's life. The preconditions are (P1) Substrate, the LLM as a software component, framed as tools, then system, then messages under prompt-caching; and (P2) Building blocks: function calling, MCP, CLI orchestration, the liteshell pattern, the agent loop, skills, characters, hooks, and scaffolding. The practices are (P3) prototype with a general-purpose agent; (P4) harvest, fold, and ship the result as a CLI, the Turtle pattern; and (P5) agent-tests-agent, in which a general-purpose agent drives it through behavioural scenarios, a complement to classical testing, not a replacement. The working loop is P3 to P4 to P5 and back, and one corollary falls out for free: multi-agent orchestration is just CLI composition. The methodology is framework-free by construction. It was distilled from the AAC, a custom agent for the open-source LAMB platform, built in about ten days by one developer with an AI pair-programmer and in production . We present it as a transferable practice, independent of any language or framework.

De novo peptide sequencing from tandem mass spectrometry is pivotal in proteomics, enabling identification of novel peptides without reference databases. While recent Transformer-based encoder-decoder models have achieved remarkable performance, we uncover a critical pathology in their inference dynamics. Through comprehensive feature scaling experiments, we demonstrate that existing auto-regressive peptide decoders tend to over-rely on generated-sequence priors while progressively under-utilizing fine-grained physical evidence from the input mass spectrum. This phenomenon leads to suboptimal results, where generated peptide sequences are biologically plausible yet not faithful to the input spectrum. To rectify this, we propose MemNovo, a training-free and plug-and-play mechanism that re-balances peptide and spectral contributions at inference time. MemNovo alleviates the information bottleneck by establishing a persistent spectral memory bank and injecting retrieved features directly into the final decoding stage via an ultra-conservative residual connection. Theoretical analysis confirms that this mechanism restores the mutual information between the decoder state and the raw spectrum. Extensive experiments on the Nine Species benchmark with two representative baselines, Casanovo and InstaNovo, demonstrate that MemNovo consistently improves both amino acid precision and peptide precision, achieving up to 39.1% relative improvement in peptide precision for Casanovo and up to 3.9% for InstaNovo, with negligible computational overhead.

In this paper, we introduce Representation Prediction via Autoencoding using Iterative Refinement (RePAIR) - a novel self-supervised representation learning architecture that synthesizes Masked Autoencoders (MAE), Joint Embedding Predictive Architectures (JEPA), and Bidirectional Encoder Representations from Transformers (BERT). We demonstrate how it can be used to encode objects in sequential data like consecutive chess positions into compact yet meaningful representations. The basic principle of the architecture is to mask large portions of a sequence of latent states, similar to BERT and MAE. Then, we apply a lightweight Predictor to the latent representations that repairs gaps in the sequence in a lower-dimensional embedding space akin to JEPA. Our experiments in the domain of chess show that the Encoder refines the board representations such that meaningful chess concepts emerge clustered in the latent space. Furthermore, reconstructions of the masked board states show that the model is able to reason about the piece movements without relying on costly reinforcement learning methods. Lastly, we find that the resulting representation space allows for quick and intuitive dissections of chess games by observing the game path trajectories in this semantically rich space.

This paper investigates an analytical Koopman-based nonlinear model predictive control (K-NMPC) approach for tracking control of virtually coupled train systems. A nonlinear train movement model incorporating train dynamics, speed and control input limits, passenger comfort constraints, and collision avoidance is systematically lifted into a finite-dimensional Koopman space through closed-form observable functions. After freezing the affine parameter-varying lifted predictor along the shifted predicted trajectory, the online optimal control problem is solved as a quadratic program that can be solved efficiently. The proposed KNMPC is benchmarked against a time-discrete NMPC scheme, demonstrating comparable control performance with significantly reduced online computation time and strong potential for real-time implementation in practical virtually coupled train control systems.

There are two main Parameter-Efficient Fine-Tuning (PEFT) techniques for Large Language Models (LLMs). While Low-Rank Adaptation (LoRA) introduces additional weights between the LLM layers, Soft Prompting introduces additional fine-tuning-specific raw tokens to an LLM input. However, both require modification to the computational graphs of precompiled, preoptimized LLMs. As a result, neither is fully supported in high-throughput engines like vLLM. We propose fine-tuning with ART (Art-based Reinforcement Training). The method injects information into a frozen Multimodal Large Language Model (MLLM) by optimizing only its raw visual input, thus enabling the soft-token approach on pre-compiled computational graphs. It relies on backpropagation of gradients back into a plain pixel array and thus supports any fine-tuning objective. Moreover, the optimized visual input can be stylized as task-relevant computational artworks. The approach's effectiveness is confirmed for different sizes of a popular open Qwen architecture and for several textual benchmarks. Specifically, ART reaches accuracy competitive with LoRA across mathematics and structured-tool-use benchmarks.

Multi-modal large language models (MLLMs) depend on in-context learning (ICL) for rapid task adaptation, but their scalability is severely limited by finite context windows and the growing cost of key-value (KV) caches in long multi-modal sequences. Existing memory compression approaches typically rely on rigid token removal or sample-dependent importance estimation, which introduces bias, disrupts semantic structure, particularly for visual representations, and yields static memories that cannot adapt to new queries. We introduce TASM (Task-Aware Structured Memory), a training-free framework that addresses these limitations through task-aware, structure-preserving, and dynamically accessible memory construction. TASM employs task-vector guided compression to replace sample-specific signals with a task-level direction that captures shared relevance across demonstrations. To preserve the underlying manifold, it applies semantics-aware token merging via bipartite graph matching, aggregating tokens without destructive pruning. Finally, TASM structures memory into a hierarchy comprising a compact Core Memory and a Latent Bank, facilitating query-adaptive dynamic retrieval. Evaluations confirm TASM maintains high performance under heavy compression, effectively balancing efficiency with adaptability.

Open-ended scientific discovery asks agents to move beyond executing analyses for predefined questions. Across multiple rounds of exploration, a discovery agent must decide which phenomena warrant investigation while avoiding overinterpretation, where emerging claims exceed the evidential scope of the analyses supporting them. This creates an evidence-calibration problem: the exploration trajectory must be coupled with claim status so that evidence can guide both what to investigate next and what can be claimed. We introduce StatefulDiscovery, a discovery framework that externalizes investigation state and uses it to coordinate frontier selection, evidence acquisition, and claim adjudication. We evaluate StatefulDiscovery across 40 real-data discovery tasks. Compared with several baselines, StatefulDiscovery produces more claims overall judged to be both well-supported and high-value. Ablations indicate that structured hypotheses, local adjudication, and frontier control contribute to performance. Together, these results suggest that explicit discovery state can couple exploration with evidence-calibrated claim formation.

Continual anomaly detection in tabular data is challenging and remains largely underexplored, particularly in settings with heterogeneous feature schemas, distribution shifts, and severe class imbalance. In many real-world applications, data arrive sequentially from diverse domains, rendering conventional continual learning methods ineffective due to their reliance on a fixed input space. We propose a continual learning (CL) method, which can overcome these challenges and continually learn from different tasks. Our method consists of three main parts: our AGF model, Taskfusion augmentation, and outlier exposure. The AGF-model maps task-specific features into a shared space, then aligns distributions to reduce representation drift, and learns anomaly decision boundaries in the aligned space. To improve stability, we introduce Taskfusion augmentation, combining boundary-aware interpolation within tasks to refine the model anomaly boundaries and cross-task mixing to transfer anomaly structure across datasets. To handle class imbalance and memory constraints, we employ tabular dataset distillation to store compact synthetic replay samples, which are jointly used with augmented data in an outlier exposure objective for robust anomaly detection. We evaluate the approach on 21 heterogeneous datasets across multiple domains. Results show that our approach substantially improves continual anomaly detection performance over sequential fine-tuning and other CL baselines while reducing catastrophic forgetting and maintaining stable detection across heterogeneous datasets.

Low-light novel view synthesis is challenging because dark multi-view images contain noise, weak structural detail, and compressed dynamic range. Recent 3D Gaussian Splatting (3DGS) methods address these challenges by generating pseudo ground-truth (pseudo-GT) images as supervision targets when paired normal-light references are unavailable. Existing pseudo-GT methods apply a uniform linear gain to all pixels, which clips bright regions while providing insufficient enhancement in dark regions, limiting reconstruction quality. We observe that nonlinear tone mappings, long established in 2D low-light enhancement, have not been explored for pseudo-GT generation in 3D reconstruction. Accordingly, we propose a scene-adaptive nonlinear tone-curve framework that replaces linear pseudo-GT with nonlinear alternatives. The framework introduces percentile-based normalisation for scene-agnostic curve application, a scene-adaptive offset for automatic black-level adjustment, and two complementary curves: Adaptive SoftExp (ASE), a bounded exponential curve, and Adaptive Poly3 (AP3), a data-driven cubic polynomial. The module changes only the pseudo-GT computation and leaves the 3DGS backbone unchanged. Experiments on three benchmarks covering 21 scenes show that both curves consistently outperform the linear baseline with PSNR improvements up to +4.34 dB on LOM and +3.25 dB on RealX3D. Both curves achieve similar performance despite their different mathematical forms, suggesting the improvement is curve-agnostic. Code is available at https://github.com/lvmingzhe/adaptiveToneCurve

Reward models for text-to-video (T2V) generation guide post-training but often fail at fine-grained semantic alignment. We trace this to two structural weaknesses in existing reasoning-based reward models: they do not systematically verify every condition described in the prompt, and the visual evidence supporting each judgment remains implicit in their free-form reasoning. We propose SG-PVR, a video reward model that addresses these limitations through plan-and-verify reasoning grounded in spatio-temporal scene graphs. The verification plan decomposes the prompt into atomic claims, ensuring every requirement is checked. The spatio-temporal scene graph, encoding entities, attributes, and temporally-grounded relations, is extracted from the video and maintained as a persistent structured visual reference throughout reasoning. Each claim is verified against both the video and the scene graph, anchoring judgments in explicit visual evidence. SG-PVR achieves strong performance on semantic alignment, including fine-grained temporal semantics. As a test-time reranker, it further enhances compositional alignment in T2V generation.

Collecting participants' lived experiences is central to design research. Focus groups are uniquely valuable because participants not only share individual accounts but also respond to one another, surfacing comparison, disagreement, and collective sensemaking. However, focus groups are resource-intensive and highly sensitive to facilitation: moderators must probe for specificity, balance participation, manage topic flow, and sustain psychological safety, and subtle facilitation choices can shape what becomes salient. Recent HCI work and commercial meeting tools show that generative AI can scaffold live conversation through prompting, turn regulation, thematic mapping, and real-time summarization. Yet UXR teams lack a clear map of what these capabilities mean in focus groups and what methodological risks they introduce. We synthesize AI supports for live conversation and translate them into a focus-group-specific playbook organized by AI role (tool, co-host, host) and modality (text, voice, embodied).We synthesize prior work on AI-supported live conversation and propose a focus-group-specific playbook of AI supports organized by role (tool, co-host, host) and modality (text, voice, embodied). We characterize interactional trade-offs and identify open questions for evaluating AI-supported focus groups as methodological configurations.

Neural relational inference (NRI) methods discover interaction graphs from trajectories through variational reasoning on discrete potential edges. However, these methods typically rely on oversimplified, factorized graph priors. Such priors, typically nearing uniform distributions, treat edges as independent entities. This systemic misalignment does not match the real-world systems and yields diffuse and indecisive edge posteriors limiting the reliability of structural discovery. To address this, we propose \textit{Diff-prior}, a diffusion-parameterized adaptive prior used to calibrate latent graph distribution rather than generate graphs. Our core insight is to reframe prior integration as a learnable denoising-style calibration that organizes scattered, uncertain edge posteriors into a more reliable overall structure which can be trained by the diffusion model. Diff-prior learns an adaptive structure prior that performs structured calibration on the edge posteriors during inference, guiding it towards a distribution closer to the underlying structure. The diff-prior operates before structural sampling and acts as a denoising calibrator directly on the encoder edge distribution, which provides a generic training paradigm over structured variables. Experiments on standard benchmarks validated our framework, and the results indicate that Diff-prior improves the performance of structure inference and generates more decisive edge posteriors across multiple NRI-family architectures. The code is available on https://github.com/Hardy158118/Diffprior.

Background. Large language models and AI agents are increasingly used to support biomedical research, but native model outputs may omit key analytical steps, misuse methods, or overstate conclusions. We evaluated whether autonomous access to a medical research skill package was associated with higher-quality AI-generated transcriptomic research-analysis outputs compared with native AI without skills. Methods. We conducted an exploratory multi-model human evaluation using a non-small cell lung cancer immunotherapy biomarker task. Six model backbones were tested. The evaluation included 21 anonymized outputs: 9 native-AI outputs and 12 skill-augmented outputs generated through an AI agent implementation represented by OpenClaw. Four non-expert biomedical reviewers and two blinded experts evaluated each output, with two ratings from each reviewer type. The primary outcome was expert-rated overall quality. Results. Skill-augmented outputs showed directionally higher expert overall quality than native-AI outputs (mean 5.50 vs 5.11; difference=0.39; bootstrap 95\% CI, -0.04 to 0.90; Welch p=0.156). Non-expert reviewer quality showed the same direction (mean 4.72 vs 4.47; difference=0.26; bootstrap 95\% CI, -0.25 to 0.80; Welch p=0.373). Expert agreement was limited (single-rating ICC=-0.15), and model-specific effects were descriptive and heterogeneous. Conclusions. Autonomous skill access showed a directional quality signal in this exploratory sample, but the signal was smaller than expert-rating noise and should not be interpreted as confirmatory evidence. The findings primarily motivate larger evaluations of skill-augmented AI agents with stronger reliability controls, platform replication, and biological-validity assessment.

Designing anthropomorphic dexterous robotic hands remains challenging as the design space straddles morphology, actuation, and sensing properties, and performance metrics span both task-dependent and task-agnostic. Existing optimization methods are often unstructured or consider only a single performance metric, limiting systematic comparison and targeted refinement. While the design considerations of the entire hand are significant, the individual finger properties play a key role in dexterity. By developing a robotic hand platform where fingers can be modularly integrated into a full teleoperated hand, we propose that optimizing the fingers can significantly improve overall hand performance. This approach enables rapid screening of different finger-level prototypes through a number of quantitative benchmarks before their integration into the hand for task-level validation. Candidate finger designs (incorporating variations in joint, bone, skin, and sensor placement) are assessed using both mechanism-oriented and task-relevant metrics, which establish a quantitative link between component design and full hand embodiment. The framework is validated through the development of an anthropomorphic robotic hand with optimized fingers, demonstrating how these fingers enable performance improvements across tasks, including multi-object grasping and light bulb screwing.

The handling of flexible materials is a difficult task to fully automate due to the challenges caused by the deformability of these types of objects. Meanwhile, a fully manual process can be ergonomically challenging, tedious and inefficient. Thus, human-robot collaboration (HRC) and cooperative manipulation (co-manipulation) have received increasing interest in this field as they enable human involvement when needed while also improving productivity. To enable efficient co-manipulation and interaction between the human operator and the robot, different modalities and control methods are required. In this paper, we present and examine different control methods for co-manipulation of carbon fiber plies, evaluating the pros and cons of each method in a controlled setting. We propose that a multimodal combination of speech commands, wrist-tracking through vision, and force with compliant control would provide the best solution for complete and intuitive control of the task.

Large Language Models (LLMs) are increasingly used for code generation, raising concerns that they may be misused to produce malicious code. Meanwhile, Grammar-Constrained Decoding (GCD) has been widely adopted to improve the reliability of LLM-generated code by enforcing syntactic validity. In this paper, we reveal a counterintuitive risk: this reliability-oriented technique can itself become an attack surface. We uncover a new jailbreak attack, termed CodeSpear, that exploits GCD to induce LLMs into generating malicious code. Our experiments show that simply applying a benign code grammar constraint can effectively jailbreak LLMs. To address this vulnerability, we propose CodeShield, a safety alignment approach that robustly preserves safe behavior even under attacker-controlled grammar constraints. CodeShield aligns the model in the code modality by teaching it to generate honeypot code under GCD. Such code is semantically harmless, so it does not implement the malicious request, and structurally diverse, so it is difficult to suppress through grammar tightening. At the same time, CodeShield still preserves natural-language refusals when natural language is available. Experiments on 10 popular LLMs across 4 benchmarks show that CodeSpear outperforms representative jailbreak baselines and increases the attack success rate by more than 30 percentage points on average. CodeShield also restores safety under CodeSpear while preserving benign utility. Our findings reveal a fundamental risk of GCD and call for greater attention to its potential security implications.

Forecasting real-world events requires language-model agents to reason under uncertainty from incomplete, time-bounded information. Yet evaluating whether agents genuinely forecast requires more than final-answer accuracy: a model may be correct by recalling memorized training facts, citing fabricated evidence, or producing an unsupported causal story. We present WorldReasoner, an evaluation framework for temporally valid event forecasting. Each task gives an agent a resolved forecasting question, a simulated forecast date, and access only to evidence available before that date; after resolution, the framework scores the submitted probability, cited evidence, and optional causal event graph. WorldReasoner reports three complementary axes: outcome quality against resolved answers, evidence quality over cited sources, and reasoning quality against post-resolution hindsight graphs. The benchmark is built by an agentic construction pipeline that generates forecasting questions, collects time-stamped evidence, and builds hindsight reference graphs at scale, yielding 345 resolved tasks derived from 14,141 articles with graphs covering 8,087 extracted events. Across six controlled agent settings, temporally valid retrieval is the strongest driver of outcome accuracy; causal graph construction improves key-event recovery; and correct graph-enabled forecasts are more strongly grounded in key events and relevant sources, yet agents still struggle to convert grounded evidence into calibrated probabilities.

Machine-learning approaches to quantum state tomography can achieve high reconstruction fidelity, but the physical structure used by the trained model often remains implicit. Here we ask whether a sparsified Kolmogorov-Arnold Network (KAN) can be used not only as a regressor, but also as an inspectable reconstruction rule whose internal organization can be checked against known Pauli structure. We study a controlled three-qubit GHZ-family benchmark in which all 63 non-identity Pauli expectation values are used to reconstruct three GHZ-subspace variables: the population imbalance $z$, the real off-diagonal component $c$, and the imaginary off-diagonal component $s$. Under finite-shot sampling and depolarizing noise, external ablation identifies the extended 12-channel GHZ-relevant Pauli set from the 63 measurements, with exact top-12 recovery across the tested shot counts and depolarizing-noise strengths. These support patterns remain stable across multi-seed random-initialization and noise-level analyses, and collapse under random-label controls. The dominant pruned input-hidden-output pathways organize Z-type population observables and X/Y off-diagonal observables in a pattern consistent with the analytic GHZ Pauli grouping, and sparse formula recovery recovers the canonical signed Pauli relations. The contribution of the KAN is therefore pathway-level structural interpretability within a neural reconstruction model, rather than superior sparse regression. Together with negative controls, these probes provide a consistency chain for auditing learned reconstruction rules against known physical structure.

Production LLM systems accumulate reusable operational experience, but the practical deployment issue is not merely whether such experience can help. It is how different serving strategies trade off quality against online cost under realistic constraints. Injecting external experience can improve task quality, yet it also increases prompt burden, latency, and serving pressure. We study \textit{external experience serving} as a deployment-oriented quality-cost trade-off problem. We evaluate this question in a real production moderation setting, with tool-use and GPQA as supporting contrast tasks that expose different output-cost regimes. We compare no-experience baselines, random experience controls, global prompt injection, and retrieval-based selective injection, and analyze both task quality and serving cost. The results show that, once experience becomes case-dependent, selective retrieval provides a stronger operating point than unconditional global injection. They further show that retrieval quality matters more than simply increasing Top-$K$, and that the same serving policy can exhibit substantially different cost-benefit profiles across short-output and decode-heavy regimes. These findings suggest that external experience is best treated as a selective, cost-aware serving decision rather than as a universal add-on. Overall, in the settings studied here, external experience pays off only when both the serving interface and the task-specific cost structure make its quality gains worth the online cost.

Trustworthy AI requires reliable data-processing pipelines, not only robust downstream predictive models. As an upstream component, data summarization determines which information is retained and passed to subsequent learning or decision modules. Therefore, adversarial perturbations to the summarization process can compromise trustworthy AI in an upstream manner: they may alter the selected summary, reduce its representativeness, and further degrade the utility of subsequent learning tasks. In this paper, we study adversarial attacks on continuous data summarization under similarity-level perturbations through DR-submodular optimization. We show that a class of multi-resolution image summarization objectives can be formulated as multilinear extensions of non-negative submodular set functions and satisfy DR-submodularity with $m$-weak monotonicity. We then formulate multi-target attack generation as a min-max problem, where one admissible perturbation of the similarity structure is optimized to degrade multiple target summarization models. To mitigate such perturbations, we formulate robust defense against mixed attack types as a regularized max-min problem. For both problems, we develop approximation algorithms with theoretical guarantees. Experiments on real-data and controlled clustered benchmarks show that the proposed attack is effective in representative low-to-moderate budget regimes and can induce downstream task-performance loss. The proposed defense improves the robustness--mitigation trade-off in structured settings, while also revealing the parameter sensitivity of robust protection on real data.

Recent progress in video generation has shown impressive visual synthesis capabilities. However, open-domain customized video generation remains limited by the lack of large-scale, annotated datasets capturing diverse identity-specific attributes. To address this, we introduce PexelsCustom-1M, the first publicly available million-scale dataset for identity-preserving video generation, containing one million curated <identity, text, video> triplets across 8,000+ categories. Leveraging this, we propose CustoMDiT, a parameter-efficient framework that adapts a pretrained multimodal Diffusion Transformer into a customized video generator with only 8% additional learnable parameters. Our method surpasses prior state-of-the-art. However, benchmarks such as DreamBooth cover only 100 classes, which is insufficient for real-world applications. To overcome this, we construct OpenCustom, a new benchmark with 1,000+ categories, created via cross-dataset knowledge fusion from ImageNet and MS-COCO. Extensive experiments confirm the advantages of both our dataset and model. We will open-source the entire ecosystem--including dataset, pipeline, benchmark, and implementations--to support further research.

Spatial reasoning remains a challenge for Multimodal Large Language Models (MLLMs), as it requires reliable multi-hop inference over both intermediate states and state transitions. Current studies often leave intermediate states unverified and treat state transitions as implicit processes, which limits reliability in multi-hop spatial reasoning. To address this, we propose State-aware Visualization-of-Thought (SVoT), a reinforcement learning framework that generates interleaved, verifiable intermediate states and visualizations. SVoT integrates transition reasoning chains into the generation processes, enabling the model to verify action preconditions and effects through interleaved textual and visual reasoning. We train SVoT via Group Relative Policy Optimization (GRPO), instantiating verification through reward design and evaluating the efficacy of different fine-grained rewards. As existing benchmarks reduce state transitions to single-variable updates, substantially simplifying the problems, we establish five domains by extending classical environments and introducing two novel domains, Pacman and Gather, that require multi-object interactions and numerical reasoning. These domains support systematic evaluation of multi-hop spatial reasoning with quantitative verification of generated intermediate states and transition reasoning. SVoT with transition-aware supervision achieves state-of-the-art performance across the introduced domains, yielding up to a 65% absolute accuracy gain on out-of-distribution test sets.

Blind grasping with a dexterous hand is a crucial manipulation capability. Nevertheless, learning such tactile-only policies for real robots remains challenging due to the tactile sim-to-real gap and the limited expressiveness of sparse tactile signals. To bridge this gap, we propose a framework for tactile-only blind grasping that is deployable on a physical multi-fingered robotic hand. Our approach combines three key components. First, we introduce a Real2Sim tactile calibration pipeline that constructs a contact-calibrated digital-twin simulator capable of reproducing real tactile signals. Second, we improve the expressiveness of sparse tactile observations using a layout-aware tactile encoder, which incorporates sensor-geometry priors through self-supervised pretraining. Third, to improve generalization to unseen objects, we train object-specific reinforcement-learning experts in the calibrated simulator and aggregate their successful grasp trajectories into a tactile-conditioned Diffusion Policy. We evaluate our method on a physical LEAP Hand equipped with distributed tactile sensing across 10 seen and 10 unseen objects. The deployed policy achieves a 27\% real-world grasp success rate across all 20 objects, without real-world grasping demonstrations or visual input. Simulation ablations show that layout-aware tactile pretraining improves grasping performance, while sensing-level evaluations confirm that Real2Sim calibration increases the consistency of tactile contact events between simulation and hardware. Together, these results suggest that contact-event calibration, geometry-aware tactile representation learning, and diffusion-based policy aggregation provide an effective path toward tactile-only blind grasping on real dexterous robotic hands. Project page:Dex-Blind-Grasp.github.io.

Large language models (LLMs) have achieved remarkable progress in language understanding, reasoning, and generation, sparking growing interest in their creative potential. Realizing this potential requires systematic and scalable methods for evaluating creativity across diverse tasks. However, most existing creativity metrics are tightly coupled to specific tasks, embedding domain assumptions into the evaluation process, and limiting scalability and generality. To address this gap, we introduce an automated, domain-agnostic framework for quantifying LLM creativity across open-ended tasks. Our approach separates the measurement apparatus from the creative task itself, enabling scalable, task-agnostic assessment. Divergent creativity is measured using semantic entropy, a reference-free and robust metric for novelty and diversity, validated against human annotations, LLM-based novelty judgments and baseline diversity measures. Convergent creativity is assessed via a novel retrieval-based multi-agent judge framework that delivers context-sensitive evaluation of task fulfilment with over 60% improved efficiency. We validate our framework in three qualitatively distinct domains: problem-solving (MacGyver), research ideation (HypoGen), and creative writing (BookMIA), using a broad suite of LLMs. Empirical results show that our framework reliably captures key facets of creativity, including novelty, diversity, and task fulfilment, and reveal how model properties, such as size, temperature, recency, and reasoning, impact creative performance. Our work establishes a reproducible and generalizable standard for automated LLM creativity evaluation, paving the way for scalable benchmarking and accelerating progress in creative AI.

Multi-turn image editing is essential for iterative design, yet current models often struggle with identity drift and error accumulation over successive steps. While existing research leverages video priors for consistency, their reliance on bidirectional attention is fundamentally misaligned with the causal, sequential nature of interactive editing. In this paper, we propose AnchorEdit, the first autoregressive (AR) diffusion-based framework designed specifically for high-resolution, long-term multi-turn editing. AnchorEdit bridges the gap between video priors and causal inference through a three-stage training curriculum: identity-preserving sing-turn pretraining, causal AR forcing fine-tuning with a novel self-rollout strategy to mitigate exposure bias, and consistency distillation for efficient 4-step generation. During inference, we introduce a memory mechanism to anchor the initial subject identity and ensure stable extrapolation across extended editing trajectories. To evaluate performance, we provide a new high-resolution multi-turn editing benchmark designed to stress-test long-horizon stability. Extensive experiments demonstrate that AnchorEdit achieves state-of-the-art results, maintaining exceptional subject fidelity and instruction following even over 10+ interaction rounds.

Large language models are now widely used for everyday learning, but the underlying interactions are typically unstructured chats rather than following a curriculum. Unlike formal online learning systems, these interactions carry no prior record of the student, so any estimate of what the student already knows must be inferred from the dialogue itself. We show that this gap is not closed by scaling models alone. Frontier and education-tuned LLMs perform poorly when asked to tutor a student over an extended session, because doing so requires three things at once. The tutor must sequence a curriculum, conduct Socratic dialogue, and infer the student's knowledge state from that dialogue. We propose separating these responsibilities. Given a student query, our system constructs a prerequisite knowledge graph in which subtopics are nodes and dependencies are edges, and frames tutoring as deciding which node to teach next and how many dialogue turns to spend on it before moving on. A lightweight PPO policy handles this sequencing decision, while an LLM conducts the Socratic exchange at the chosen node and returns a signal of student progress. Across held-out STEM and non-STEM topics, our PPO-paired tutor outperforms heuristic baselines, frontier general-purpose models, and a model specialised for Socratic dialogue: on both the rate at which students reach full curriculum mastery and the number of turns required. Explicit curriculum structure delivers gains that scaling the underlying model does not.

Vision-language-action (VLA) models provide strong visual, language, and action priors for robot manipulation, but visual observations alone often miss the local contact state required for contact-rich tasks. We present TacCoRL, a scalable framework that injects Tactile feedback into VLA policies and improves them through sim-real Co-training and simulation-based reinforcement learning (RL), without requiring large-scale tactile pretraining or extensive real-world contact exploration. The key idea is not only adding touch as an input, but learning how contact readings should modulate action responses in near-failure states that are rare in demonstrations and risky to collect on hardware. We use a real-aligned simulator as a closed-loop training environment for contact interaction. Mixed simulated and real trajectories first warm-start tactile-conditioned actions in the pretrained policy. Reinforcement learning with verifiable task rewards then optimizes the policy using simulated contact rollouts. It reinforces tactile-conditioned actions that lead to task completion, while a supervised objective on real trajectories keeps the refined policy anchored to deployment visual, tactile, and action distributions. The resulting policy transfers directly to the real robot without privileged simulation state or online real-world RL. Across four bimanual contact-rich tasks, the final visuo-tactile policy achieves an average success rate of 72.5%, compared to baseline of 50.0%. Result videos and more details are available at https://tac-corl.github.io/

We introduce a multi-view in-cabin monitoring dataset for public transportation with synchronized RGB and depth images from four inward-facing cameras and a rotating LiDAR covering the vehicle interior of a digitalized and partly automated German city bus. The dataset contains 9.136 synchronized samples with annotations and is accompanied by a calibration and pseudo-labeling pipeline that generates 3D human pose estimates and oriented 3D bounding boxes for occupants. We further provide a nuScenes-format conversion and benchmark representative multi-view 3D detection models (e.g., Lift-Splat-Shoot and BEVFusion), supporting comparative evaluation and small-scale training of multi-view in-cabin perception models. The dataset and tools are available at https://github.com/EvgenyGorelik/multiview_incabin_dataset.

Spatial reasoning remains a persistent challenge for multimodal large language models (MLLMs). Existing approaches largely rely on large-scale, statically curated datasets, where all training samples are treated uniformly regardless of the model's evolving capabilities. This static paradigm is inherently data-inefficient: training capacity is often spent on samples that are either trivial or overly difficult for the model at its current stage. To address this limitation, we propose Ouroboros-Spatial, a self-evolving training framework in which the model plays dual roles as a proposer and a solver. In each iteration, a frozen proposer generates spatial question-answer (QA) pairs from 3D scene metadata and raw video frames, together with executable code for deriving reliable ground truth. A learnable solver is then fine-tuned on the accepted samples, and its per-sample prediction confidence is used as a difficulty signal. This signal is fed back to the proposer in the next iteration, guiding it to generate questions better matched to the solver's current capabilities. Through this closed-loop design, the training distribution co-evolves with model ability, reducing redundant trivial examples while filtering out ambiguous or uninformative samples with limited learning value. Across six spatial reasoning benchmarks, Ouroboros-Spatial substantially improves Qwen3-VL-4B and Qwen3-VL-8B while using an order of magnitude fewer training examples than recent large-scale curated datasets. On VSI-Bench, it yields absolute gains of 9.9 and 6.8 points for the 4B and 8B models, respectively, enabling both to outperform a wide range of strong open-source and proprietary baselines.

User-side memory in LLMs is typically scored as a single "personalization" capability: given a user's history, is the output more user-aware? We show this aggregate metric hides opposite-direction failures. Memory factorises into at least three orthogonal axes -- behavioral consistency (style, voice), factual presence (recall facts in history), and factual absence (abstain when a fact is absent) -- and no single substrate wins all three. Comparing per-user gamma-LoRA (a small LoRA adapter trained on each user's history; gamma denotes per-user, not per-task) against BGE-large dense top-K retrieval on a controlled 50-user synthetic corpus and a real-data probe (LaMP-3), we find gamma-LoRA decisively wins behavioral style while RAG decisively wins factual absence -- and the same query-projection cells in attention layers 21-35 causally load-bear both effects in opposite directions (zeroing those LoRA weights raises absence-probe TPR by +33 pp and drops presence-probe TPR by 20 pp). On the more heavily RLHF-tuned Llama-3.1-8B-Instruct the asymmetry strengthens, not heals: parametric memory's behavioral advantage collapses while its absence-calibration deficit against retrieval widens -- an alignment tax on parametric user-memory. On real-data LaMP-3, gamma-LoRA underperforms a majority baseline; a 9-condition mitigation sweep diagnoses this as instruction-following collapse, not substrate failure (a 9x2 cross-product shows the eval-time {1..5} logit mask drives main_acc to >=0.995 on every recipe), and the best training-time fix replicates bit-identically on Llama. Finally, substrate-selection routing is question-classification, not calibration: a 110M DistilBERT on the question text alone beats every logit-based router. We contribute the diagnostic framework, the diagnosed real-data negative, the alignment-tax replication, and the routing-as-classification finding.

Online learning with delayed feedback typically assumes that the learner can track all pending rounds until their feedback arrives. In practice, tracking resources are finite, and feedback from untracked rounds is permanently lost. In this paper, we study delayed online convex optimization (OCO) under a hard capacity constraint, where at most $C$ pending rounds can be tracked at any time. To model delay information, we introduce a semi-clairvoyant model that refines the clairvoyant assumption from prior work: rather than requiring delays to be known at prediction time, the learner observes delay expirations online, consistent with the classical unconstrained delayed setting. Our approach proceeds via a reduction to a novel ``delayed and weighted'' OCO problem, using a scheduler that randomizes tracking decisions and importance-weights the resulting observations. For this base problem, we propose and analyze Delayed-Weighted FTRL and its bandit analogue, establishing regret bounds that explicitly characterize the interaction between time-varying weights and delayed feedback. Combining these base learners with our schedulers yields the first regret guarantees for capacity-constrained OCO under convex and strongly convex losses, for both first-order and bandit feedback. For first-order feedback, capacity $C = Ω(\log T)$ suffices to recover standard delayed OCO rates up to logarithmic factors. For bandit feedback, the regret rates are modulated by powers of $(1 + σ_{\text{max}}/C)$, where $σ_{\text{max}}$ is the maximum number of pending observations at any time. This allows the regret bound to degrade gracefully when $C < σ_{\text{max}}$, while remaining sublinear.

This paper presents ERN-Net, an Evolving Reason Node-Net for efficient document image binarization. ERN-Net enhances degradation-sensitive regions, such as faint strokes, broken characters, and noisy backgrounds, through evolving reason nodes and multi-scale reasoning. We further compare ResNet-101, ConvNeXt-Tiny, and ConvNeXt-Base, and find that ConvNeXt-Tiny provides the best practical trade-off between accuracy and memory usage. In addition, DIBCO-based pretraining improves binarization performance without increasing model memory consumption, requiring only about 1.5 additional training hours. Experiments on DIBCO-style benchmarks show that ERN-Net is effective under low-data and low-memory settings.

Autonomous exploration with UAVs in large-scale, topologically complex environments often suffers from low efficiency due to suboptimal scheduling and detours. Prior maps (e.g., construction drawings), although usually imprecise and flawed, are readily available in many scenarios and have the potential to provide global structural guidance. This paper presents a novel exploration framework that leverages sparse, unaligned, and even discrepant 2D prior maps for LiDAR-based UAV exploration. First, a robust 2D-3D point cloud registration pipeline is proposed to align LiDAR observations with prior maps. The registration pipeline combines a GeoContext descriptor for single-frame candidate retrieval, a multi-frame verification mechanism for coarse transformation estimation with outlier rejection, and a Scale-ICP algorithm for refinement. The registration module can handle map discrepancies and provide multiple hypotheses when geometric ambiguities arise. To effectively utilize the registration results for exploration planning, we further develop a hierarchical viewpoint planning strategy under localization uncertainties. The hierarchical strategy first spatially attaches local viewpoints to prior guidepoints and adopts a Monte Carlo Tree Search solver to determine their traversal sequence under each registration hypothesis. To mitigate registration uncertainty, a risk-aware selector evaluates prior sequences using confidence-weighted travel risk, and a fixed-endpoint traveling salesman problem is formulated to generate an efficient local coverage path under the selected prior guidance. Benchmark evaluations reveal up to 34.2% improvement in exploration efficiency and 37.9% reduction in flight distance compared to state-of-the-art methods, while extensive simulations and field experiments further demonstrate robustness to prior map incompleteness and deformations.

To make clinically grounded decisions, medical AI agents are expected to go beyond simple recognition and be capable of tool retrieval, evidence acquisition, and integration. Existing benchmarks largely evaluate isolated perception or single-turn question answering, and therefore provide limited visibility into failures of planning, tool recruitment, and rollout reliability. We introduce MedCTA, a benchmark for evaluating medical tool agents on clinician-validated, step-implicit tasks grounded in realistic multimodal clinical inputs, including radiology images, pathology slides, and reports. MedCTA comprises 107 real-world clinical tasks with clinician-verified executable trajectories over 5 deployed tools, and supports process-aware evaluation of tool selection, argument validity, execution stability, trajectory fidelity, and outcome quality. We benchmark 18 open- and closed-source multimodal models and find that even frontier systems remain brittle in multi-step clinical tool use: autonomous rollouts are dominated by protocol failures, premature stopping, and incorrect tool recruitment, while gold-standard tool routing yields large but still incomplete gains. These results show that strong backbone perception does not translate into reliable agentic behavior in clinical settings. MedCTA provides a rigorous testbed for auditing, diagnosing, and advancing trustworthy medical AI agents. The dataset and evaluation suite are available at https://ivul-kaust.github.io/MedCTA/

Deliberative polling promises to improve collective decision-making by exposing shareholders to a broad range of arguments before they vote. Yet ensuring that every voter encounters a representative sample of the reason space, the coverage problem, remains an open challenge, particularly at scale and in adversarial or strategically motivated electorates. This paper introduces a way of evaluating solutions using the LLM-based Agentic Bipolar Argumentation Simulator, grounded in a framework which formalises a poll as a six-tuple <Jend, Jopp, Ratt, Renh, VA, VR> of endorsing and opposing justifications, attack and enhance relations, and shareholder- and relation-weights. ABAS simulates N autonomous shareholder agents, each assigned a latent opinion according to desired distributions in [-1, 1], who sequentially vote, choose or author justifications, and optionally submit argumentation-graph links. The simulator implements recommendations that rank existing justifications by their observable endorsement mass. It evaluates the mechanism's success by coverage, namely the fraction of the corpus reason-tag set represented in the K recommendations presented to each shareholder, as a solution to the NP-hard Subsuming Justification Problem. Reported experiments characterise how creativity rate (pown), recommendation size (K), argumentation density (plinks), and population size (N) affect coverage and corpus diversity. In an authenticated electorate where Sybil attacks are impossible and only the relation graph is gameable, we stress-test the scoring with coordinated strategic voting attacks: a tag-flood attack collapses coverage, while author-count relation weighting through a reversed-PageRank rule resists the flood markedly better than uniform weights.

Composed Image Retrieval (CIR) constitutes a pivotal paradigm requiring models to perform joint reasoning on reference images and modification texts. However, the prevalence of Noisy Triplet Correspondence (NTC) in large-scale datasets severely constrains model performance. Existing denoising methods either target binary mismatches or rely on scalar-based point-wise estimation, neglecting rich global structural correlations among sample populations and dynamic value variations during training, thereby yielding suboptimal results. This paper identifies two critical unresolved challenges: Global Structural Inconsistency of Semantic Correlations and Hard Sample Discrimination Uncertainty. To address these, we propose RankVR, a framework designed to construct a robust CIR model via global structure consistency and dynamic value perception. Specifically, we introduce the Global Structure Consistency Perception (GSCP) module, which utilizes the Effective Rank of the Correlation Matrix to decouple clean samples from structural noise. By measuring rank difference, GSCP identifies samples disrupting macroscopic semantic symmetry. Furthermore, we develop the Adaptive Semantic Value Calibration (ASVC) module to distinguish high-value hard clean samples. By integrating training potential and reliability, it dynamically quantifies the semantic value of each triplet, ensuring effective utilization of hard samples while suppressing noise characterized by logical conflicts. Extensive experiments on the FashionIQ and CIRR benchmark datasets demonstrate that RankVR significantly outperforms existing state-of-the-art methods, validating its superior robustness in noisy environments.

Long-horizon LLM agents are not trusted to run unattended: with no human watching, they confidently report success they never verified. We treat honesty -- bounding what an agent may claim at termination -- as a first-class metric for unattended autonomy, distinct from capability. We present Autopilot, an execution model that makes silent fabricated success structurally impossible rather than merely rarer. Autopilot externalizes all working state into a durable, gated finite-state machine that a scheduler advances one stateless tick at a time; a hard floor forbids any terminal "done" claim whose falsifiable gate did not actually execute and pass. We prove a No-False-Success theorem -- under gate soundness, floor enforcement, and plan coverage, termination implies the goal holds -- whose only trust points are empirically measurable, and show the worst case degrades to an honest stall, never a fabricated success. Because each tick rehydrates only the state machine, per-step context cost is constant in the horizon. Across a 3,150-cell paired corpus (70 tasks $\times$ 3 systems $\times$ 3 models $\times$ 5 seeds, including 50 SWE-bench Lite tasks across 11 OSS repos), Autopilot fabricates on 0.95% of cells [95% CI 0.38--1.62] while Reflexion and StateFlow baselines fabricate on 8.10% [6.48--9.81] and 25.05% [22.48--27.62] respectively. The headline contrast lives in the hard regime: on SWE-bench Lite, the firewall reduces fabrication from 33.7% (StateFlow) to 0.67%, a paired difference of $-33.07$ pp [95% CI $-36.53, -29.73$]. The mechanism is the gate, not the model: all ten Autopilot fabrications come from the strongest model, while two weaker mid-tier models never fabricate across 700 paired cells. The firewall trades coverage for honesty by design -- an honest stall is recoverable; a confident wrong output shipped downstream is not.

End-to-end task-success is the dominant way to evaluate LLM agents, but one aggregate number tells you that an agent regressed, not where. We present layer-isolated evaluation: a deployed ordering agent is decomposed into a fixed taxonomy of layers (ontology, intent, routing, decomposition, escalation, safety, memory, and cross-cutting envelope/defense), each exercised by its own assertion slice in a deterministic, no-LLM "pure" mode. The pure suite (238 cases across 23 slices; 225 run in 2.39 s, ~10 ms/case) runs in CI on every change against a locked per-slice baseline. We validate by controlled regression injection, degrading one layer at a time across seven non-safety layers. The effect we did not design in is masking: the aggregate pass-rate barely moves (-1.7 to -5.9 pp for six local regressions), while the matching slice craters (-25 to -91 pp). A layer's slice reacting to its own fault is partly by construction; the measured results are (i) the aggregate masking and (ii) that damage stays off the other slices: the injected layer's slice is the single worst-hit in 5 of 7 cases and top-3 in 7 of 7 (mean rank 1.29 of 19). Localization replicates on a second, structurally different tenant (Starbucks SG): all seven matching slices crater, so it is not a single-catalog artifact. We position it as a concrete, deterministic instantiation of the component-level evaluation EDDOps prescribes but leaves unimplemented, with CheckList as ancestor and as the deterministic mirror image of whole-workflow stochastic mutation testing. Our contributions: (a) a fully decomposed, sub-second, no-LLM per-layer harness for a production agent, (b) a coverage-honesty test-adequacy criterion that refuses to score an unexercised layer, and (c) the regression-injection demonstration that per-slice baseline-locked gates localize regressions an aggregate metric masks.

Large language model (LLM) agents struggle with long-horizon tasks due to their inherent statelessness, requiring all task-relevant information to be encoded in growing input contexts. The resulting degraded reasoning quality, increased inference cost, and higher latency necessitate efficient working memory mechanisms. However, existing approaches either rely on lossy compression or similarity-based retrieval, which often fail to capture temporal structure and causal dependencies required for multi-step agentic tasks. In this work, we present HORMA, a Hierarchical Organize-and-Retrieve Memory Agent that organizes experience into a file-system-like hierarchical structure, where summarized entities are linked to the corresponding raw trajectories, enabling efficient access without losing detailed information. HORMA decomposes working memory into two stages: structured memory construction and navigation-based retrieval. The construction module iteratively refines how experiences are structured by distinguishing between failures caused by missing information and those caused by misleading or overloaded context. The navigation module retrieves task-relevant context by traversing the hierarchy using a lightweight agent trained with reinforcement learning to select minimal yet sufficient context, thereby reducing latency along the critical execution path. Across ALFWorld, LoCoMo, and LongMemEval, HORMA improves task performance under constrained context budgets while requiring at most 22.17% of the baseline token usage in long conversation tasks. Compared to existing methods, it consistently achieves better efficiency-performance trade-offs and generalizes effectively to unseen tasks.

Problem, Research Strategy, and Findings: The rise of large language models (LLMs) raises a key question for urban planning: which forms of professional planning knowledge can AI replicate, and which still require human judgment? Although AI tools are increasingly used in planning practice, there is still no systematic framework for testing whether they can reason with the contextual sensitivity, value awareness, and institutional literacy central to planning expertise. This paper introduces Urban Planning Bench (UPBench), a domain-specific evaluation framework that assesses LLM reasoning through a 4x5 matrix of four knowledge pillars and five cognitive levels adapted from Bloom's revised taxonomy. Evaluating 25 LLMs with automated scoring and expert review, we find a non-monotonic cognitive curve: models perform better on higher-order analytical tasks than on factual recall and integrative judgment. This suggests that planning knowledge often treated as lower-order is deeply shaped by institutional, jurisdictional, and temporal context, making it hard for LLMs to generalize. We summarize these limits as four epistemic diagnostics: regulatory hallucination, conceptual conflation, wickedness paralysis, and phronetic deficit. Takeaway for Practice: The findings support differential delegation in planning. LLMs can assist with cross-disciplinary synthesis, literature review, scenario generation, and preliminary policy analysis. However, they remain unreliable for jurisdiction-specific regulation, normative conflict resolution, and context-sensitive procedure. Agencies should require verification for AI-assisted regulatory analysis, while planning education should emphasize institutional literacy, normative judgment, and contextual sensitivity.

Subject-preserving video generation is not solved by frontal-face similarity alone: a generated person must remain recognizable across motion, large viewpoint changes, expression shifts, occlusion, scale variation, and conflicts among text, first-frame, and identity references. We argue that the central bottleneck is the point-reference paradigm, which collapses identity into a single static observation entangled with pose, accessories, lighting, background, and camera statistics. We introduce Argus, a Wan-based framework centered on Stacked Multi-View Identity Mosaic Injection (SMII). SMII converts MLLM-selected image/video identity evidence into a 3*3 stacked mosaic, synchronizes the mosaic with the current diffusion time, and injects it as negative-time read-only memory in Wan's native token space. This turns identity from an external clean adapter or a single reference image into a compact dynamic distribution. Around SMII, an MLLM Identity Director selects informative identity moments and resolves condition conflicts, while no-cross-pair counterfactual training, Temporal Identity Annealing, and Adaptive Self-Likeness Guidance improve robustness without paired subject-video supervision. We further release HardID-Celeb, a public-figure identity-stress benchmark, and introduce YawScore and OccScore to probe large-yaw and first-frame-occlusion robustness. Argus achieves state-of-the-art results on OpenS2V-Eval Human-Domain, reaching 64.38 Total Score, 71.86 FaceSim, 51.62 NexusScore, and 79.14 NaturalScore. On HardID-Celeb, Argus obtains 76.80 FaceSim and improves YawScore and OccScore by 12.60 and 15.10 points over the strongest baselines, demonstrating that dynamic identity memory and large-scale counterfactual self-supervision are highly effective for subject-preserving video generation.

Agentic infrastructure introduces a critical control-plane authorization problem: non-deterministic reasoning systems can propose high-stakes mutations to production resources, yet existing security mechanisms -- such as identity and access management (IAM), policy engines, consensus protocols, and audit logs -- either enforce static, context-unaware permissions or merely record actions post-execution. This paper introduces the Sovereign Assurance Boundary (SAB), a certificate-bound runtime admission layer for autonomous execution authority. SAB intercepts agent proposals at an assurance airlock, compiles them into typed execution contracts $C$, and binds these contracts to cryptographic evidence digests $H(E)$ and policy versions. The contracts are then routed through consequence-aware certification paths. Upon successful admission, the system emits a signed Sovereign Assurance Certificate ($Ω$) that is strictly scoped to a specific execution identity, revocation epoch, and validity window. Finally, a sovereign execution broker verifies $Ω$ and performs fresh pre-execution revocation and drift checks before invoking infrastructure APIs. We detail the airlock-broker architecture, formalize its admission and revocation invariants, and report preliminary feasibility measurements from a Go prototype evaluated over 2,500 admission attempts. Ultimately, this broker-enforced model prevents autonomous reasoning from directly mutating state, transforming delegated execution authority into a cryptographically verifiable, evidence-bound, revocable, and replayable runtime artifact.

The most widely-adopted robot learning pipelines today learn skills from robot demonstrations or structured human data, which are expensive to collect and tied to specific embodiments. In contrast, unstructured human videos provide a scalable alternative. They contain diverse manipulation demonstrations across objects, scenes, and strategies, but are not directly connected to robot action. We propose LUCID, a two-stage framework that learns task intent from unstructured human videos drawn from internet-scale datasets and learns robot control in massively-parallel simulation. The intent model predicts short-horizon intent (what should happen next in the scene) from the current observation in closed loop. An embodiment-specific sensorimotor policy converts this intent into robot actions. The intent interface is shared across controllers, so the same intent model can be applied to different embodiments, from our primary dexterous hand to a parallel-jaw gripper. We evaluate LUCID on five real-world manipulation tasks: stirring, wiping, and binning supervised by only internet video, with zero-shot transfer to novel scenes and object instances; and push-T and cable routing supervised by 1 hr each of self-collected smartphone video. Project page: https://lucid-robot.github.io/.

The widespread adoption of face recognition (FR) technologies raises serious privacy concerns, as facial data can be exploited without consent. To address this challenge, we propose Adv-TGD, a generative adversarial attack framework that synthesizes photorealistic faces capable of impersonating target identities and deceiving face recognition systems. Built upon Stable Diffusion, Adv-TGD performs per-sample LoRA fine-tuning conditioned on concise textual prompts to generate natural yet adversarially manipulated identities. Unlike conventional identity-attack approaches, our method optimizes lightweight cross-attention adapters for each source-target pair within a single-step denoising process. Latent blending is constrained by a face-local heatmap mask to ensure spatially precise identity manipulation while preserving non-sensitive regions. We introduce a composite objective that integrates masked epsilon-MSE reconstruction, thresholded identity divergence in FR embedding space, directional feature alignment, and source-similarity suppression to balance adversarial attack and visual realism. Optionally, LLaVA-generated attribute prompts enhance fine-grained semantic details without reintroducing identity cues. Under the black-box evaluation protocol, Adv-TGD attains an average attack success rate (ASR) of 85.90% across IR152, IRSE50, MobileFace, and FaceNet, surpassing the semantic SOTA baseline Adv-CPG by +6.25 points, diffusion-based makeup method DiffAIM by +3 points, and noise-based P3-Mask by +16 points. Despite its strong attack efficacy, Adv-TGD preserves high visual fidelity (PSNR = 27.15 dB, SSIM = 0.981). Furthermore, we demonstrate the flexibility of our framework by successfully extending it to in-the-wild datasets (LADN), general object classification (ImageNet), and transformer-based diffusion models (FLUX.1).

Stance detection requires identifying an author's position toward a target, often from short-form texts where stance is implicit, indirect, or rhetorically framed. Although large language models (LLMs) achieve strong performance on this task, single-pass prompting can be brittle when multiple interpretations are plausible. Existing aggregation strategies, such as majority voting or self-consistency, improve robustness by combining labels, but they discard the intermediate reasoning needed to resolve conflicting interpretations. We introduce a multi-agent reasoning framework with adaptive worker allocation for stance detection that shifts aggregation from label-level voting to reasoning-level synthesis. The framework employs a Manager-Worker architecture in which a Manager adaptively allocates a variable number of Worker agents based on input complexity. Each Worker analyzes the input from a distinct perspective and produces a reasoning-only explanation without emitting a stance label; the Manager then synthesizes these explanations to produce the final prediction. We evaluate the proposed framework on SemEval-2016, P-Stance, and COVID-19 Stance using Llama, Mistral, and Gemini. Results show that the framework yields the largest gains on implicit and context-dependent stance cases, achieving 86.07 Macro-F1 on COVID-19 and 82.90 on SemEval-2016, while remaining competitive on more explicit stance datasets such as P-Stance. These findings suggest that adaptive reasoning-level aggregation is most beneficial when stance cannot be reliably inferred from surface cues alone.

Audio-visual Generalized Zero-shot Learning (AV-GZSL) is a challenging task that aims to classify both seen and unseen objects or scenes by integrating data from audio and visual modalities. Recent studies primarily focus on fusing or aligning audio and visual features to generate more informative audio-visual embeddings. Also, aligning the audio-visual and textual features of most existing methods relies solely on the optimization objectives. However, those methods neglect the inherent distributional and structural differences between audio-visual and textual modalities. To address this limitation, we propose a method termed Aligning Hierarchical Standardized Embedding (AHSE), which enables hierarchical alignment of standardized audio-visual and textual embeddings within a shared embedding space. Specifically, we first apply Z-score standardization to the fused audio-visual and textual embeddings to reduce distributional mismatches. We then introduce a hierarchical alignment strategy that minimizes discrepancies at the semantic, class, and batch levels, thereby constructing a more robust and well-structured embedding space. This strategy not only preserves semantic and inter-class relationships but also maintains spatial consistency within each batch. Extensive experiments on three benchmark datasets: VGGSound-GZSL, UCF-GZSL, and ActivityNet-GZSL, demonstrate that AHSE achieves competitive performance in zero-shot learning.

The rapid development of intelligent control methodologies has endowed robots with powerful autonomous intelligence. Cable routing, a ubiquitous foundational task in industry, provides a rigorous benchmark for robotic dexterity and sequential decision-making. In these practical scenarios, image observation distortion frequently occurs. Samples characterized by low-quality image observations often hinder accurate model training, posing challenges to the reliability and accuracy of intelligent control systems. Nevertheless, no dedicated intelligent control solution has been proposed for scenarios of image signal distortion. Meanwhile, image quality information has not been sufficiently exploited to further enhance the performance of intelligent control methodologies. To this end, we propose a novel robotic imitation learning framework that comprises an image quality assessment module, a confidence-based learning mechanism, and a decision-making module, which is designed to maintain high performance even under distorted image observations. In the proposed framework, the image quality assessment module synergizes with the confidence-based learning mechanism to enhance the efficacy of the decision-making module. Specifically, the image quality assessment module is incorporated to extract image quality information from image observations, while the confidence-based learning mechanism adaptively prioritizes challenging samples to improve learning effectiveness. The decision-making module determines appropriate discrete skills or continuous actions. Experimental results demonstrate that our formulated framework enhances the overall performance of the decision-making module.

Modern Vision-Language Models (VLMs) benefit from chain-of-thought prompting and test-time scaling, but these gains often come with prohibitive inference cost due to large visual contexts and long decoding chains. We view this cost through two coupled axes: Visual Context Scaling (VCS), which controls how much visual evidence is passed to the language model, and Visual Reasoning Scaling (VRS), which controls how much inference-time reasoning search is performed. Existing methods typically optimize one axis at a time, leaving the joint allocation of compute across these axes underexplored. We introduce Adaptive Visual Inference Scaling (AVIS), a lightweight policy that adapts both VCS and VRS per query. AVIS realizes VCS through Key Diversity Visual (KDV) pruning, a training-free $O(N)$ key-based rule for removing redundant visual tokens before prefilling, and realizes VRS through adaptive self-consistency, using a learned difficulty predictor to select the number of reasoning rollouts. AVIS is deployment-friendly and compatible with shared-prefill inference, where all rollouts reuse a single prefilling pass and KV cache. Across diverse image and video reasoning benchmarks, AVIS improves the accuracy--compute trade-off relative to VCS-only and VRS-only baselines, and remains effective on top of RL post-trained VLMs while keeping compute and latency low.

Despite recent advances, Vision Language Models (VLMs) still struggle to grasp the dynamics of the world. We note that the ability to reason about a 4D scene, challenging in itself, is further complicated by two factors. First, VLMs observe motion indirectly via its projection onto 2D images. Second, existing datasets fail to disentangle object and camera motion. To address these challenges, we present a QA generation pipeline that focuses on motion-related scene understanding. We take particular care of the entanglement of camera and object motion by casting tracking in both the traditional way and in a novel, fixed reference system, dubbed True-Motion Tracking, which provides an intuitive description of motion. From this pipeline, we generate a large-scale training dataset of 400K samples, 4DP-QA (4D Perception QA), and a 2.2K-sample benchmark, 4DP-QA-Bench. Training existing models on our dataset yields performance improvements on an external benchmark, validating the effectiveness of our method.

Natural environments present a complex challenge to robotics perception systems. Current models, particularly vision foundation models, are largely trained on structured, urban environments leading to weaknesses in their perception for field robotics tasks. We showcase the limitations of current models using our recently released WildCross benchmark, a new cross-modal benchmark for place recognition and metric depth estimation in large-scale natural environments. WildCross comprises over 476K sequential RGB frames with semi-dense depth and surface normal annotations, each aligned with accurate 6DoF pose and synchronized dense lidar submaps. In this work, we provide an expanded analysis of the benchmark results from the recent WildCross benchmark, with particular emphasis on expanded metric depth estimation experiments. Access to the code repository and dataset for this work can be found at https://csiro-robotics.github.io/WildCross.

Graph analysis underlies many applications whose answers cannot be looked up in a single record or retrieved along a path: laundering rings, drug repurposing, user preference, and scientific theme are all inferred from a node together with its neighbourhood. We introduce GraphInfer-Bench, a benchmark for whether LLMs can perform this graph inference: producing an open-ended answer that no single node supports and no path retrieves. Existing graph-QA protocols cannot test this capability: algorithm simulation, node classification, single-node description, KG-QA, and GraphRAG all admit answers retrievable from one node or along a path. GraphInfer-Bench defines five tasks along Description (what a region is) and Comparison (how regions differ), each constructed so the ground truth lives in no single node. The release contains 42,000 samples across six real-world graphs, produced automatically and screened by a four-layer quality-control protocol. We evaluate four method families against the same tasks: graph-token alignment models, zero-shot frontier closed-source LLMs, Graph2Text supervised fine-tuning, and plain GNNs as a structural reference. No method family closes the gap. Graph-token alignment partially handles description tasks (relational, theme) but collapses on comparison tasks. Frontier LLMs lead on outlier detection and community partition among LLM-based methods but lag on masked-node prediction. Graph2Text SFT is the strongest LLM-based method on the description side yet falls behind frontier LLMs on comparison. Across every task, plain GNNs match or beat the strongest LLM-based row, with the largest margin on community detection. GraphInfer-Bench surfaces graph inference as an open capability gap rather than a property of any one architecture.

Large language models (LLMs) achieve remarkable performance across a wide range of tasks, but their autoregressive decoding process incurs substantial inference costs due to inherently sequential token generation. Speculative decoding addresses this bottleneck by employing a lightweight draft model to propose multiple future tokens that are subsequently verified in parallel by a larger target model. Recent work has demonstrated that diffusion language models are well suited for this setting, as they can generate entire blocks of draft tokens in parallel and thereby alleviate the sequential constraints of autoregressive drafting. A subtlety of this regime is that block-diffusion drafters generate tokens bidirectionally within a block, whereas verification is performed by an autoregressive target model that evaluates tokens in a strictly left-to-right manner, leaving a gap between the symmetric training-time objective and the asymmetric verification-time reward. In this work, we offer an empirical analysis of three training-time interventions that narrow this gap: token positional weighting, a first-error focal loss that targets the position that breaks the accepted prefix within each block, and a chain loss term that substitutes a differentiable surrogate for the expected accepted length. The three interventions act along orthogonal axes (position, block-conditional first error, joint prefix) and compose additively; they are likewise orthogonal to test-time alignment mechanisms such as multi-draft self-selection, with which they can in principle be combined. Across four target models and six reasoning, code, and dialogue benchmarks, the three interventions raise accepted draft length by 21-76% per benchmark over a position-uniform baseline, without adding additional forward passes and without changing the inference pipeline or the rejection-sampling exactness contract.

Vision-language models like CLIP can provide rich semantic priors for open-vocabulary object detection. However, jointly integrating both textual and visual knowledge into detection architectures remains challenging. In this paper, we propose VL-DINO, an open-vocabulary detector that enhances DINO through more effective exploitation of CLIP's vision-language knowledge. Specifically, a Query-guided Positive Sample Construction (QPSC) module is first developed to construct additional high-quality positive samples, enabling the vanilla DINO framework to better accommodate mixed training across heterogeneous data sources while providing more vision-language alignment signals, thereby incorporating richer textual knowledge during training. A Visual Semantic Encoder (VSE) module is then introduced to distill CLIP visual knowledge into backbone-extracted features, producing fused features for subsequent encoder refinement. Based on the fused features, an Object-Region Semantic Alignment (ORSA) module extracts object-centric region features and aligns them with the corresponding textual embeddings, further incorporating textual cues. In the zero-shot setting, VL-DINO-T and VL-DINO-L achieve 36.3 and 38.1 AP on the LVIS benchmark, respectively, consistently outperforming prior advanced approaches. Extensive experiments demonstrate the effectiveness and competitive performance of the proposed design.

Learning-based policies are being considered to augment the dexterity of human surgeons in robot-assisted surgery. Can the end-to-end mapping from visual observations to robot actions be vulnerable to adversarial attacks, potentially leading to patient injury? In this paper, we present the first study of adversarial threats to learning-based policies in surgical robotics. We investigate two threat modes: (a) disruptive attacks, where imperceptible visual perturbations interrupt policy execution, and (b) steering attacks, where such perturbations steer policy actions toward attacker-specified directions. We formulate three adversarial attack methods, each with increasing access to policy information, and evaluate their impact on two surgical subtasks: debridement and suturing. Our evaluation covers three end-to-end policy architectures: ACT, Diffusion Policy, and Pi0. In addition, we introduce a new class of photometric adversarial attacks that mimic natural visual changes, such as lighting variations, to generate effective yet visually plausible perturbations. Results from 560 physical experiments using phantoms for debridement and suturing suggest that state-of-the-art policies can be significantly disrupted, resulting in an average 61% reduction in surgical subtask success rates. Project page: https://sites.google.com/view/adversary-surgery

We introduce the ladderpath index as a measure of language complexity grounded in algorithmic information theory. It counts the minimum steps needed to reconstruct a sequence through hierarchical reuse of repeated substructures, capturing an exactly computable but constrained form of algorithmic compressibility related to, but distinct from, Kolmogorov complexity. We apply the ladderpath approach to 21 parallel corpora from the Parallel Universal Dependencies dataset. The ladderpath index is approximately invariant across the languages, and varies much less than the corpus length. This is more pronounced when all corpora are mapped to a unified binary representation, providing evidence for the equi-complexity hypothesis from a representation-independent perspective. We also observe trade-offs between character inventory size and corpus length, and between vocabulary-level and corpus-level reconstruction complexity, supporting the trade-off hypothesis that total complexity is conserved and redistributed across linguistic levels. The reusable substructures identified by the ladderpath approach, without any linguistic input, overlap with words and morphological components attested in the natural vocabulary. The hierarchical reuse captured by the ladderpath approach parallels the chunking mechanisms proposed in cognitive science, where the human cognitive system compresses linguistic input into nested, reusable units under shared memory and processing constraints. This connection between cognitive chunking and the ladderpath approach provides a new interpretation for the equi-complexity and trade-off hypotheses, grounding both in the shared cognitive architecture that underlies language processing across human languages.

Contrastive Reinforcement Learning (CRL) has seen recent success in a wide variety of goal-conditioned robotics tasks by learning structured representations of the dynamics. However, despite its success in locomotion and simpler control domains, CRL often struggles in interaction-rich manipulation. We argue that a key source of this difficulty is object-centric interaction, such as contact or grasping, that induces distinct changes in the underlying dynamic modes. In this work, we formulate manipulation dynamics as a piecewise-smooth Markov process and show that interaction-induced mode changes create piecewise nonlinear reachability structures that are difficult for standard CRL energy functions to represent and plan over. Based on this analysis, we introduce Interaction-weighted Resampling (IWR). IWR performs interaction-aware resampling around phases before, during, and after interactions, encouraging the learned representation to preserve the mode boundaries that determine future reachability to capture multi-modal and piecewise nonlinear reachability. Across interaction-centric environments, including 2D dynamic control, robotic manipulation, and robot air hockey, IWR improves both sample efficiency and overall performance over prior CRL methods, with 19.8% average improvement in simulation. Finally, using a sim-to-real pipeline with policies trained by IWR, we demonstrate the first real-world goal-conditioned robot air hockey agent capable of hitting goals, improving success from 25% to 60%. Project Page: IWR-arxiv.github.io.

Training capable OS agents requires data that simultaneously captures structured user intents, multi-turn task delegation, and grounded tool execution--properties absent from existing datasets. We propose ISE (Intent -> Simulate -> Execute), a three-stage synthesis paradigm that addresses these gaps jointly. Stage 1 constructs roughly 50000 structured intents via a 4D framework (Persona x Domain x Task x Complexity); after deduplication the pool contains 43956 unique intents and attains a Vendi Score of 61.57 over the entire pool on mpnet-base-v2 embeddings (cosine kernel, q=1). Stage 2 drives multi-turn user-agent interaction through a role-locked user simulator that grounds each user turn in actual execution outcomes, producing 23132 complete trajectories averaging 8.12 user turns and 68.24 total dialogue turns. Stage 3 runs every tool call inside a live, isolated OS workspace, generating authentic failure-recovery dynamics instead of simulated responses. Fine-tuning on ISETrace improves ClawEval pass@1 from 19.3 to 37.7 using Qwen3-8B on agent tool-use tasks with a standard protocol. This result outperforms zero-shot GPT-4o and the larger Qwen3-32B base model which is four times bigger. An ablation on Stage 2 proves multi-turn simulation brings a large portion of the performance gain. We release all source code and dataset at https://github.com/Valiere01/ISE-Trace.

Indoor localization from wireless measurements remains challenging in large-scale deployments due to substantial variation in building geometry, the set of detectable access points (APs), and the heterogeneity of received signals. Existing learning-based methods often perform well only in limited settings and degrade under environmental shifts, making robust anchor-free localization across diverse indoor environments notoriously difficult. In this paper, we present OmniLoc, an environment-interactive foundation model for anchor-free user equipment localization across diverse indoor environments. To the best of our knowledge, OmniLoc is the first foundation-model-based approach built directly on wireless measurements for this task. OmniLoc is built on three key designs. First, a unified input tokenization module converts heterogeneous wireless measurements into a common representation that is more amenable to learning. Second, a geometry-aware Transformer performs AP-aware feature extraction by emphasizing dominant APs while aggregating complementary evidence from supporting APs. Third, a geometry-aware location estimation module conditions regression on geometric embeddings to produce geometrically consistent location predictions. We evaluate OmniLoc on both a large-scale in-house dataset and a public benchmark dataset. Results show that OmniLoc significantly outperforms existing methods, consistently improves existing backbones when its design components are integrated, and demonstrates strong generalization in cross-environment evaluations.

Real-world robots need to adapt their behavior beyond the envelope of their pre-trained policy. Policy finetuning or retraining are options, but they risk catastrophic forgetting, degrading the pretrained policy's base performance. To combat this, we introduce CLAE: Closed-Loop Affine Activation Editing, an inference-time framework for steering the behavior of a frozen policy by editing intermediate activations while keeping the base policy weights and downstream action head untouched. CLAE approaches behavior steering as a closed-loop problem whose outputs edit policy activations that adapt online to the robot state, environment, target behavior, and multi-robot context. It trains a sparse autoencoder over frozen-policy activations, selects behavior-relevant latent features via post-hoc probing, and learns a lightweight RL-based steering policy that applies state-dependent affine edits to selected latents during inference. We validate CLAE on a frozen multi-quadrotor navigation policy trained to perform a single task: navigating robots to a set of goal locations while avoiding obstacles. Through extensive simulations and physical tests, we show that while navigating to their goal positions, CLAE can 1. steer individual robot behavior by controlling each robot's velocity profile; 2. coordinate multirobot behavior by preserving a desired formation; and 3. produce entirely new behavior wherein robots are required to reduce their exposure to surveillance cameras in the environment.

Understanding and predicting how social beliefs evolve in response to events -- from policy changes to scientific breakthroughs -- remains a fundamental challenge in social science. Given LLMs' commonsense knowledge and social intelligence, we ask: Can LLMs model the dynamics of social beliefs following social events? In this work, we introduce the concept of the Social World Model (SWM), a general framework designed to capture how social beliefs evolve in response to major events. SWM learns state-transition functions for social beliefs by mining temporal patterns in social data and optimizing the evidence lower bound, without the need for explicit human annotations linking events to belief shifts, or for expensive census data. To evaluate SWM, we introduce a benchmark, SWM-bench, derived from real-world prediction markets, specifically Kalshi and Polymarket. SWM-bench includes over 12k data points for social belief prediction tasks spanning diverse domains such as politics, finance, and cryptocurrency. Our experimental results show that SWM significantly outperforms time-series foundation models, achieving state-of-the-art results on Kalshi data and demonstrating competitive performance on Polymarket data, while offering interpretable insights into the underlying mechanisms of social belief dynamics.

Large Language Models (LLMs) have achieved strong performance across natural language processing tasks, yet reliable reasoning remains an open challenge. Although modern LLMs show progress in structured inference, multi-step problem solving, and contextual understanding, their reasoning behavior is often inconsistent and sensitive to prompting strategies, task design, and model scale. This survey provides a systematic analysis of more than 300 recent papers from arXiv, Semantic Scholar, Google Scholar, Papers with Code, and the ACL Anthology to examine how reasoning capabilities emerge in LLMs and where they fail. We make three main contributions. First, we introduce a structured taxonomy of LLM reasoning research, covering Chain-of-Thought reasoning, multi-hop reasoning, mathematical reasoning, common sense reasoning, visual and temporal reasoning, code and algorithmic reasoning, retrieval-augmented reasoning, tool-augmented and agentic reasoning, and reinforcement learning-based reasoning. Second, we analyze methodological trends across these paradigms, including prompting methods, model architectures, training objectives, reward modeling, and evaluation benchmarks. Third, we synthesize recurring limitations and failure modes, such as reasoning hallucinations, brittle multi-step inference, weak causal abstraction, and poor cross-domain generalization. By organizing a rapidly expanding literature, this survey offers a unified view of the current capabilities and limitations of reasoning in LLMs. We also identify emerging research directions, including meta-reasoning, self-evolving reasoning frameworks, multimodal reasoning, and socially grounded reasoning. Overall, this work aims to serve as a reference for developing more robust, interpretable, and generalizable reasoning systems in future language models.

Robotic table tennis is a representative benchmark for high-speed, closed-loop robotic control in dynamic environments, where accurate and fast prediction of ball states is critical for reliable planning and control. Physics-based approaches rely heavily on accurate parameter identification and precise initial state, while learning-based methods often struggle to capture long-range temporal dependencies and are typically trained on limited or simulated data. We propose a transformer-based framework for table tennis ball state prediction that leverages attention mechanisms to model long-range temporal correlations directly from historical observations, without relying on explicit flight or bounce models. To support robust learning and generalization, we collected a large-scale real-world dataset from players of varying skill levels and diverse ball cannon configurations. The combination of a high-capacity transformer architecture and extensive real-world data enables accurate long-horizon forecasting. Building on this capability, we introduce a plug-and-play sim-to-real transfer strategy, Swap Predictor at Deployment (SPAD), which replaces the physics-based simulator used during training with the proposed real-world-trained predictor at deployment, improving the sim-to-real transferability of the policy without requiring retraining. We demonstrate that this simple substitution effectively narrows the sim-to-real gap while preserving the efficiency and scalability of simulation-based training.

Large Language Models are highly sensitive to prompt formulation, necessitating automatic prompt optimization to unlock their full potential. While evolutionary algorithms have emerged as the dominant paradigm, they suffer from a critical bottleneck: data efficiency. Current methods treat the development dataset as a static benchmark, wasting significant compute budget on uninformative data. In this work, we introduce APEX (Automatic Prompt Engineering eXpert), a novel framework that optimizes the data usage alongside the prompt search. APEX dynamically stratifies the dataset into Easy, Hard, and Mixed tiers based on the optimization lineage. By prioritizing the Mixed tier, which identifies the data where the LLM has mixed performance, we identify two high-leverage subsets: the addressable frontier for generating informative mutations and the rank-sensitive frontier for distinguishing candidate quality. We evaluate APEX across three diverse benchmarks: IFBench, SimpleQA Verified, and FACTS Grounding. Under a fixed budget of 5,000 evaluation calls, due to its data efficiency, APEX outperforms the initial prompt by an average of 11.2% on Gemini 2.5 Flash and 6.8% on Gemma 3 27B, demonstrating that a data-centric approach is key to efficient and effective prompt optimization.

The deployment of LLM-based agents in scientific analysis raises opposing concerns: that agents may reduce methodological diversity, or that they may amplify the analytic flexibility through which researchers reach motivated conclusions. We argue these worries target two empirically separable layers: a design layer of methodological choices, and a verdict layer in which a decision rule maps estimates to a substantive claim. We test both by running 20 independent executions of Claude Code and Codex on a prominent immigration and social-policy against a many-analysts human baseline. At the design layer, Codex matches human methodological diversity and Claude Code produces nearly three times as many specifications; both agents' effect estimates remain broadly aligned with the human consensus, and no agent model exactly matches any human model. A prompt-induced anti-immigration researcher prior reorganizes each agent's methodological decisions but, unlike for biased human analysts in the same data, does not shift aggregate estimates or final verdicts; nor do agents reroute along the methodological axes humans use to bias their estimates. At the verdict layer, an explicit confirmatory prompt flips Claude Code's verdicts from 10% to 90% support while leaving its coefficient distribution essentially unchanged, operating through rule omission rather than rule softening. AI agents can rival or exceed human methodological diversity at the design layer while remaining vulnerable at the verdict layer. In our setting, the locus of AI bias is not estimation but interpretation.

Recent anecdotal evidence suggests that AI coding agents can reproduce published findings when provided with original data and code; yet systematic evaluation across social sciences remains limited. Existing evaluation benchmarks are insufficient, either small or conflate agent performance with problems in the reproduction materials themselves, such as code that fails to execute correctly. Here we introduce SocSci-Repro-Bench, a benchmark of 221 tasks spanning four disciplines and 13 substantive domains, constructed from studies whose results are either fully reproducible with available materials or demonstrably non-reproducible due to missing data, allowing us to isolate agents' reproduction capacity. Evaluating two frontier coding agents, Claude Code and Codex, we find that both can reproduce a large share of social science findings, with Claude Code substantially outperforming Codex. These reproduction rates considerably exceed those previously reported for general-purpose LLM-based agents on comparable reproducibility benchmarks. Both agents also perform strongly on a reasoning task requiring identification of underlying research questions, and additional analyses suggest that results are not primarily driven by memorization. Providing the original paper PDF alongside replication materials modestly improves performance but introduces bias on tasks where reproduction is impossible. We also show that agents can be nudged toward confirmatory specification search through subtle prompt framing. Together, these findings suggest that at least some frontier coding agents can serve as reliable executors of computational workflows while underscoring the need for careful benchmarking and prompt design as AI systems assume larger roles in scientific production.

The growth of agent skills has transformed how agentic systems are built, evaluated, and deployed. As skill libraries continue to scale, rigorous evaluation becomes critical to ensuring their utility, quality, and safety in real-world applications. Consequently, the field is undergoing an emerging paradigm shift from isolated skill creation to automated, evaluation-driven skill evolution. In this survey, we systematically examine the landscape of skill evolution and evaluation beyond foundational skill creation. We categorize evolution into four distinct paradigms, spanning execution feedback, trajectory distillation, compression, and reinforcement learning, showing how each element contributes to improving skill utility and reliability. We also provide an analysis of six skill-centric benchmark categories, identifying structural gaps in benchmark coverage, trade-offs, and metric richness to advance skill research. Finally, we identify open directions for building skill ecosystems that are generalizable, efficient, and verifiably safe. The project URL is https://github.com/Cassie07/AgentSkill_Survey

Natural language interfaces to databases aim to translate user questions into executable SQL, yet remain brittle in real-world settings where questions are underspecified and schemas are large and ambiguous. Ambiguity across user questions, database schemas, and model interpretations are central failure modes in NL2SQL, leading to misaligned intent, incorrect schema grounding, and erroneous SQL generation. Existing approaches rely on human clarification or treat ambiguity as a schema representation problem, but these do not scale nor resolve ambiguity autonomously. We propose SOMA-SQL to automatically resolve ambiguity via targeted synthetic query log and ambiguity-driven probing. SOMA-SQL constructs synthetic query log to ground schema interpretation and guide candidate SQL generation; it then executes targeted probing queries, driven by a structured ambiguity taxonomy and candidate disagreements, to produce disambiguation evidence for final SQL selection and repair. This active approach to ambiguity discovery and resolution generalizes across unseen schemas and query distributions without human-in-the-loop. Experiments on six public benchmarks demonstrate that SOMA-SQL improves execution accuracy by 13.0% on average over state-of-the-art baselines, with gains of up to 16.7% on ambiguous questions.

Every day, millions absorb claims from podcasts and streams that no fact-checker ever sees. Spoken misinformation is built through conversation, where credibility comes not from facts alone but from how claims are framed, reinforced, or left unchallenged across turns. Yet fact-checking has focused on isolated text, leaving dialogue audio under-studied. We introduce MAD2, a new Multi-turn Audio Dialogues benchmark for spoken claim verification, containing 1,000 two-speaker dialogues with 3,368 check-worthy claims and approximately 10 hours of audio, and propose calibrated multimodal fusion of a context-aware audio encoder and a dialogue-aware text model. Across settings, adding dialogue context improves verification, but the gains depend on scenario type. Using only preceding context often matches offline performance, supporting live-moderation settings, and audio contributes most when transcript-based models are destabilized by additional context. Overall, conversational structure matters more for verification than misinformation framing.

Action chunking has become a standard design in modern robot policies, from diffusion/flow policies to vision-language-action models, where the policy predicts a sequence of actions and executes a fixed number of them instead of acting one step at a time. However, this paradigm relies on a key assumption: a fixed execution horizon. During chunk execution, the policy operates open-loop, which is particularly problematic for fine-grained manipulation tasks that require frequent replanning. In practice, the execution horizon is typically chosen through empirical tuning and is highly task-dependent. To this end, we propose Dynamic Execution Horizon Prediction (DEHP), an effective method that trains a lightweight execution-horizon prediction branch using online reinforcement learning while keeping the pretrained chunk policy completely frozen. This makes the method compatible with black-box chunk policies and isolates the effect of adapting the execution horizon from changes to the underlying action generator. Across our evaluations, DEHP improves the success rate of different high-precision and long-horizon manipulation tasks by a large margin. Our qualitative analysis further shows that DEHP predicts shorter execution horizons during fine-grained stages of the task and longer horizons during free-space motion. In this way, DEHP balances the efficiency of open-loop chunk execution with the reactivity of closed-loop single-step control. Project page: https://dehp-chunking.github.io/

Dexterous manipulation with multi-finger hands can be sensitive to physical parameters such as object shape, pose, and friction coefficients. While simulation enables large-scale data collection with known parameter values, simulation-trained policies must still handle uncertainty at deployment, where the true parameters and therefore the true dynamics are unknown. Standard domain randomization strategies may be insufficient for precise tasks like screwdriver turning, as manipulation strategies may need to change depending on specific parameter values. To address this, we propose Probabilistic Latent Unified world Modeling and parameter Estimation (PLUME), a world model that jointly learns to evolve a belief over parameter values as well as the system dynamics conditioned on those parameters. We learn a latent space to jointly represent multiple qualitatively different physical parameters along with rewards, themselves functions of partially-observable variables, to inform planning. Our novel learning framework leads to efficient alignment of the world model to true dynamics through online parameter inference as opposed to re-training or fine-tuning. We evaluate our method on simulated screwdriver turning, valve turning, bucket lifting, and disk flicking tasks, as well as a hardware screwdriver turning task, where we achieve successful zero-shot transfer of our simulation-trained policy and outperform state-of-the-art offline reinforcement learning and world-model-augmented behavior cloning baselines. Please see our website at https://plume-world-model.github.io for videos.

Full-duplex spoken language models (FD-SLMs) enable seamless speech interaction by allowing models to listen and speak simultaneously, yet the internal mechanism by which they coordinate listening and speaking remains underexplored. We analyze the predictive behavior encoded in FD-SLM hidden representations and find that they exhibit stream-specific predictive patterns: during listening, they preferentially predict the incoming user stream, whereas during speaking, they preferentially predict the model output stream. Building on this observation, we show that FD-SLMs dynamically modulate their internal predictive focus between two states: a generative state aligned with model output generation and a perceptive state aligned with incoming user input. However, this modulation can lag behind abrupt changes in conversational context. During user interruptions, the model remains transiently biased toward the generative state before transitioning into the perceptive state, causing it to miss the beginning of the incoming input. We term this delayed internal transition state inertia. To quantify its downstream impact, we introduce the Zero-Buffer Benchmark (ZBB), a diagnostic benchmark for evaluating immediate interruption comprehension when user speech begins abruptly. We evaluate this setting using response correctness and initial-word occurrence rate (IWOR). Finally, we mitigate state inertia through activation steering with a perception vector, a training-free intervention with little additional computational overhead. Across multiple state-of-the-art FD-SLMs, activation steering substantially improves interruption handling; for example, on PersonaPlex, it improves correctness from 28% to 45% and IWOR from 40% to 72% without any fine-tuning.

Although symmetricity in the converter controller is desirable for robust stability margins, a direct link between system-level asymmetricity and instability has yet to be clearly established. Converter control introduces three-phase asymmetricity through loops such as DC-link voltage control, a phase-locked loop , and a power synchronization loop. Furthermore, the inherently asymmetric topology of the two-level voltage-source converter, which converts a DC voltage into a three-phase balanced set, acts as the underlying origin of the asymmetries that propagate into the control structure. Consequently, establishing a direct relationship between system asymmetricity (rather than control asymmetricity alone) and the stability margin is essential for understanding the underlying instability mechanisms. In this work, asymmetricity is quantified using the Asymmetricity Quantification Index (AQI), derived from the sequence-domain representation of the interconnected converter-grid impedance. Within this domain, symmetricity is identified through the definition of symmetrical matrices, which serve as the benchmark against which asymmetricity is measured. A robust and generalized analysis correlates AQI with the stability margin, including both grid-following and grid-forming control structures connected to the power grid. It is found that instability arises from increased asymmetricity in the combined converter-grid system, which is dominated by asymmetric control loops and operating points. Thus, reducing asymmetricity without compromising controller functionality can improve stability margins. The analysis is validated in both control-hardware-in-the-loop and power-hardware-in-the-loop environments.

Piezoresistive tactile sensors are attractive for robotic manipulation because they are thin, lightweight, low-cost, and scalable to dense large-area sensing. However, existing systems still face a practical trade-off: recent reproducible designs emphasize accessibility and ease of reproduction, whereas high-fidelity readout architectures remain more difficult to fabricate, assemble, and deploy. We present HiPi, a reproducible high-fidelity piezoresistive sensing system for robotic manipulation. Building on a low-crosstalk readout principle, HiPi redesigns the complete hardware stack around reproducibility, deployability, and multi-sensor scalability. The system includes a compact readout PCB compatible with commercial PCB fabrication and assembly services, eliminating manual soldering; a smaller and lower-cost STM32-based MCU module; an optimized communication pipeline that achieves 220 Hz readout in a bimanual setup with four dense tactile arrays (2048 taxels in total); and FPCB-based conductive layers that simplify sensor fabrication and stacking. Experiments with structured 3D-printed contact patterns show that HiPi preserves contact geometry substantially better than a reproducible baseline, improving the average IoU from 0.428 to 0.797 and the average Dice score from 0.539 to 0.886. These results suggest that HiPi bridges an important gap between reproducible fabrication and high-fidelity readout, making dense piezoresistive tactile sensing more practical for bimanual manipulation and multi-fingered robotic systems.

Structured abstracts are important for biomedical literature processing, by facilitating information retrieval, text mining, and knowledge synthesis. However, a vast portion of abstracts indexed in PubMed remain unstructured, presenting a significant bottleneck for downstream text-processing workflows and applications. To resolve this limitation, we introduce Structured PubMed, a comprehensive corpus of section-labeled biomedical abstracts compiled from the complete PubMed database, encompassing over 23.2 million research-article records. The corpus is divided into two distinct subsets: a collection of 5.9 million author-structured abstracts parsed from official XML files, and an automatically labeled collection of 17.2 million originally unstructured abstracts structured via a verbatim-extraction Large Language Model pipeline. Every record is harmonized under a unified five-section schema and mapped to its original PubMed identifier, publication type, and publication date. This dataset can be utilized to train sentence-classification models, benchmark text-segmentation architectures, and perform large-scale, section-specific information extraction at an unprecedented PubMed-wide scale.

Retrieval-augmented generation degrades when scaled to large, heterogeneous document collections, where dense similarity loses discriminative power, and top-k retrieval increasingly returns semantically similar but contextually incorrect chunks. We refer to this failure mode as vector search dilution. Even when using hybrid dense+sparse retrieval, we observed this firsthand in a deployed Wyoming Department of Transportation corpus, where scaling from 54 to 1,128 documents (88,907 chunks) reduced accuracy from 75% to below 40%. To address this dilution, we propose MASDR-RAG ( Multi-Agent Scoped Domain Retrieval for RAG) and evaluate it on 200 expert-validated queries across five LLM backbones, six corpora, and two index stacks. Our results indicate that domain scoping using organizational metadata is the key fix, significantly improving P@10 from 0.77 to 0.86 ($p < 0.05$). Furthermore, our investigation of multi-agent orchestration revealed that a high degree of configuration dependence results --creating what we call the precision-faithfulness paradox. Based on these varied outcomes, our practical recommendation is simple: scope first, then perform a single synthesis call, reserving full multi-agent orchestration for genuinely multi-domain corpora paired with native-tool-call backbones. Code and Data will be made public upon acceptance.

Modular neural network pipelines suffer from error compounding: noise at any module boundary propagates and potentially amplifies through subsequent modules. We introduce energy conservation as a hard physical constraint on inter-module information flow. Activation energy (the squared L2 norm of feature vectors) is enforced to be exactly preserved at every module boundary. Unlike soft energy penalties, conservation is an inviolable law: the network may redistribute energy across neurons but cannot create or destroy it. Four experiments on CIFAR-10 demonstrate: (1) conservation retains 77.4% of clean accuracy at noise sigma=0.2, versus 35.1% for baselines and 30.9% for energy-penalized models (p<0.001, 5 seeds); (2) pipelines become depth-invariant, retaining 93.3% at depths 2 through 5 with noise at every boundary; (3) the advantage generalizes to systematic bias (+45.1%), Gaussian (+40.4%), and adversarial noise (+4.8%), with a principled non-effect on dropout (-0.3%); (4) on ResNet-18, the conservation advantage scales inversely with intrinsic normalization: +0.3 pp with BatchNorm, +26.2 pp without at sigma=0.2, reaching +58.0 pp at sigma=0.5. Experiment 5 validates the operator on a real modular robotic pipeline (MuJoCo physics, Franka Panda). Across three independent runs on separate machines (90 trials per cell), conservation provides +18.9 pp average advantage on monocular-depth-style noise. A formal bound proves conserved noise energy is strictly less than input noise energy.

Scientific AI agents increasingly retrieve evidence, reason across sources, and synthesize conclusions used in consequential decisions. Yet, their ability to do so in high-stakes domains such as health remains unclear. We introduce SciConBench, a large-scale live benchmark of 9.11K questions and expert-written conclusions from systematic reviews to evaluate open-domain scientific conclusion synthesis. The benchmark draws on an expert-validated automated evaluation pipeline that decomposes conclusions into atomic facts and measures correctness and comprehensiveness via factual precision and recall. To mitigate data leakage, we further introduce SciConHarness, a clean-room evaluation harness that equips agents with controlled web interaction to ensure valid measurement. Evaluating 8 frontier models and deep research agents, we find that factual quality remains low: under clean-room settings, the best agent achieves only a factual F1 of 0.337. Our clean-room setting consistently reduces performance relative to unconstrained evaluation, suggesting that leakage inflates estimates of models' true synthesis capabilities. Finally, we audit consumer-facing agents (e.g., Google AI Overview, OpenEvidence) and find they frequently generate incomplete and sometimes contradictory conclusions, even when the ground-truth answer is available. Overall, our results show that reliable synthesis of scientific conclusions remains an open challenge, and that clean-room evaluation is essential for assessing open-domain AI agents.

Sustained driving automation systems are envisioned to be used as the foundation for driverless mobility services. However, both researchers and practitioners acknowledge that current driving automation systems are not yet able to handle all traffic situations that a human driver can handle. To bridge this gap and enable mobility services without an in-vehicle human driver or fallback, remote operation (or teleoperation) is increasingly discussed. Recently, first legal actions have been taken to enable some forms of remote operation on public roads. Remote operation encompasses a broad spectrum of methods to support a driving automation system, ranging from remote assistance, which includes providing information or releasing a maneuver, to remote driving, which includes driving the vehicle from a remote location. As such, safe implementation of remote operation in public road traffic challenges the collaboration of multiple academic disciplines (e.g. engineering, psychology, informatics, law, etc.) and stakeholders (e.g. remote operation service providers, remote operators, vehicle manufacturers, regulatory authorities, etc.). At the same time, the interdisciplinary discourse is often challenging due to differing expectations and language. To build a common ground, this article traces terminology back to the original differences in information processing both on human and vehicle side. This framework aims to help further discourse by directly specifying what is needed to engage a diverse audience including researchers and stakeholders of different backgrounds and interests. Recently discussed forms of teleoperation are integrated into this framework.

We introduce Embodied-R1.5, a unified Embodied Foundation Model (EFM) that integrates comprehensive embodied reasoning capabilities, spanning embodied cognition, task planning, correction, and pointing, within a single architecture toward general physical intelligence. Leveraging three automated data construction pipelines to significantly expand the data coverage of critical capabilities, we build a large-scale data system of over 15B tokens, and design a multi-task balanced RL recipe to alleviate heterogeneous task conflicts. We further introduce a Planner-Grounder-Corrector (PGC) closed-loop framework that enables a single model to autonomously execute and self-correct over long-horizon tasks. With only 8B parameters, Embodied-R1.5 achieves SOTA on 16 out of 24 embodied VLM benchmarks, surpassing leading models like Gemini-Robotics-ER-1.5 and GPT-5.4. Benefiting from the internalized embodied capabilities, Embodied-R1.5 can be fine-tuned into a VLA with only a small amount of data, outperforming leading VLA models like $π_{0.5}$ across 4 popular manipulation benchmark suites. We further conduct extensive zero-shot real-robot experiments, validating performance in instruction following, affordance grounding, articulated object manipulation, and long-horizon complex tasks, demonstrating strong generalization to the physical world. We open-source model weights, datasets, training code, and EmbodiedEvalKit, an evaluation framework tailored for embodied tasks, to facilitate future research in EFMs.

Contact-rich manipulation requires robots to continuously perceive and regulate evolving physical interactions under dynamic contact transitions or complex surface geometries. Recent imitation learning methods improve contact-aware control by incorporating tactile or force feedback, but they rarely model the asymmetric spatiotemporal roles of global force and local tactile sensing. To address this, we propose TacForeSight, a lightweight force-conditioned tactile foresight framework for real-time manipulation. The core component is TacForceWM, a tactile world model that predicts short-horizon tactile latent dynamics from dual-finger tactile observations conditioned on high-frequency wrist force and torque signals. Another key component, the Predictive Tactile-Conditioned Policy, leverages the predicted latents as anticipatory contact priors, models the current-to-future tactile evolution via cross-attention, and adaptively fuses visuo-tactile features through a tactile-guided gating module. By forecasting purely within a compact latent space, TacForeSight enables proactive contact reasoning with efficient real-time inference suitable for high-frequency manipulation control. Real-robot experiments on five representative tasks and three in-process perturbation settings show that TacForeSight consistently outperforms existing baselines, particularly under dynamic contact disturbances. All models and datasets will be made publicly available on the project website at https://tacforesight.github.io/ProjectPage.

Large Language Model (LLM)-based multi-agent systems are increasingly powerful, but current agentic workflow optimization paradigms make an unsatisfying trade-off. Task-level methods spend substantial offline compute yet deploy only a single workflow, leaving complementary candidates unused, while query-level methods synthesize a new workflow per query at substantial inference cost. Our motivating analysis shows these paradigms are more complementary than competing: workflows discovered during offline search often solve different subsets of queries, and many queries handled by expensive query-level generation can already be solved by cheaper precomputed workflows. This suggests a different objective: rather than searching for one universally best workflow or regenerating one per instance, we should build a compact bank of reusable, complementary workflows and select among them adaptively at inference time. Doing so requires solving three coupled problems: generating complementary rather than redundant candidates, compressing them into a small deployable portfolio, and assigning each query to the right workflow under a performance-cost trade-off. To this end, we present FlowBank, a three-stage framework for portfolio-based agentic workflow optimization. Diversifying proposes DiverseFlow to steer search toward under-covered queries and produce a high-coverage candidate pool. Curating proposes CuraFlow to compress this pool into a compact portfolio with minimal redundancy. Matching casts deployment as edge-value prediction on a query-workflow bipartite graph and routes each incoming query to the portfolio member with the best predicted utility. Across five benchmarks, FlowBank achieves the highest average score among the evaluated methods while remaining cost-competitive, improving over the strongest automated and handcrafted baselines by 4.26% and 14.92% relative, respectively.

Data tells stories that shape society; the data journalist's job is to turn raw information into stories non-experts can trust. A high-quality news feature takes a newsroom team weeks: hunting for context, running statistics, choosing an angle, and designing visuals. Recent agents handle individual steps well: data-science agents close the analysis loop, while design agents synthesize beautiful websites. But can an agent serve as a data journalist end to end? We introduce Data Journalist Agent (Data2Story), a multi-agent framework that orchestrates specialized roles into a single virtual newsroom. Data2Story contributes two innovations. (i) Claims are evidence-grounded: an Inspector links every number, angle, and asset back to data, code, or an external reference. (ii) Articles are multimodally generative: rather than defaulting to plain text and static charts, Data2Story reasons about what readers will want to see, then deploys multimodal tools, such as interactive maps for geography and audio for music. We evaluate Data2Story on 18 articles, each paired with the originally published expert piece, along four axes: (a) human-agent angle coverage; (b) rubric evaluation with 53 participants across five dimensions; (c) computer-use agents as judges, a cost-saving proxy for how readers navigate interactive articles; and (d) verifiability, where a coding verifier re-executes statements against the data and checks claims against references. Data2Story produces competitive, evidence-traceable multimedia stories, with particular strength in transparency and auditability. Human articles retain an edge in editorial angle, creative design, and presentation. We position Data2Story as a collaborator for journalists, enabling more evidence-based, transparent, and verifiable reporting. Code and demos are available at https://data2story.github.io.

Assistive mobility and manipulation platforms have received increasing attention as a means of restoring independence to individuals with disabilities. While effective for many basic activities of daily living (ADLs), a significant percentage of everyday tasks such as opening a jar, pouring a liquid, lifting a tray, or basic meal preparation, is fundamentally bimanual and remains out of reach for any single-arm system. Adding a second arm to a wheelchair is impractical, due to the additional power draw, cost, and the loss of space required for transfers and mobility. We instead propose a heterogeneous, on-demand bimanual system, in which a wheelchair-mounted anchor arm is joined when needed by a summoned mobile manipulator that serves as a complement arm. The central technical problem, which we call bimanual joining, is conditional: the anchor has already committed to a grasp, and the complement arm must choose where to stand and what to grasp to complete the task. We formulate bimanual joining as a three-phase decomposition (plan, drive, grasp) and show that a vision-language model (VLM), coupled with standard geometric tools, provides task-level knowledge sufficient to solve a representative class of bimanual ADLs. Our system JOIN, contributes (i) a wheelchair-referenced opposition score, and (ii) task-conditioned directional manipulability. We evaluate JOIN on a Kinova Gen3 anchor and a Hello Robot Stretch~3 complement on representative same-object and different-object tasks. JOIN accomplished more attempts (19/20) than state-of-the-art methods (14/20) and required markedly less correction by the operator.

Falls are one of the leading causes of injury and hospitalization among elderly individuals, making reliable fall awareness an essential capability for safety monitoring in residential environments. However, existing fall detection systems often rely on wearable devices or fixed sensing installations, which may suffer from low user compliance, limited spatial coverage, or degraded performance under occlusion and poor lighting conditions. In this work, we propose \textbf{EM-Fall}, an embodied fall detection framework deployed on a mobile humanoid robot. The system integrates millimeter-wave (mmWave) sensing with robotic mobility, allowing the robot to actively adjust its sensing viewpoint and maintain target observability across rooms and under occlusion. To address interference in complex residential environments, including pet motion and multipath artifacts, we design a human-centered perception pipeline combined with lightweight temporal modeling to capture motion evolution before, during, and after fall events. We evaluate the proposed system across eight real indoor environments with four participants and construct an in-home mmWave fall detection dataset. Experimental results show that the embodied mobile sensing paradigm improves monitoring continuity and maintains robust fall detection performance under diverse environmental conditions. The proposed framework provides a practical solution for robot-assisted safety monitoring in home environments.

Elite humanoid soccer shooting requires whole-body stability, high-impulse whole-body interactions, and accuracy to targets. Motion tracking-driven reinforcement learning (RL) provides stability in whole-body movement coordination, but a fixed reference makes it hard to adapt to varied ball positions and strike timings; in contrast, task reward-driven RL struggles to explore and discover valid kicks from scratch. We therefore introduce RoboNaldo, a three-stage motion-guided curriculum RL framework for high-impulse humanoid interaction. A single human-kick reference is used as a scaffold and progressively shifts optimization towards shooting performance. The curriculum first learns a stable whole-body kicking prior, then adapts the kick to free-kick settings where the ball is stationary at random positions, and finally extends it to moving-ball shooting through a locomotion-command and kick-trigger interface. A high-level heuristic planner controls this interface during training, while alternative high-level controllers can drive the same low-level policy at inference. In simulation, RoboNaldo demonstrates free-kick shot error 48.6% lower and shoot velocity 2.96x than prior work baselines. In real world on a Unitree G1 with onboard perception, RoboNaldo attains 0.73 m and 0.86 m average target shooting error from 3 m away in free-kick and moving-ball cases, accordingly. And the post-contact ball velocity reaches 13.10 m/s, which is 59-71% of reported professional open-play shot speed. Project page: https://opendrivelab.com/RoboNaldo.

Real-world multi-agent systems, from traffic coordination to resource allocation, are often modeled as general-sum games where individual incentives conflict with collective welfare. In these settings, the central challenge is not merely finding an equilibrium, but selecting socially desirable outcomes among many suboptimal Nash equilibria. Standard deep multi-agent reinforcement learning (MARL) methods struggle with this problem, as value-decomposition approaches are constrained by monotonicity assumptions and policy-gradient methods often converge to stable but socially inefficient equilibria. To address this limitation, we propose $Φ$-Actor-Critic ($Φ$-AC), a framework that leverages swap regret minimization to steer learning toward high-welfare correlated equilibria (CE). To make counterfactual regret estimation tractable in deep MARL, $Φ$-AC employs a centralized attention critic that predicts vector-valued regrets in a single forward pass, avoiding computationally expensive counterfactual simulations. We further introduce a Lagrangian-based equilibrium selection mechanism that optimizes social welfare while enforcing stability through regret constraints. Experiments on matrix games, Multi-Agent Particle Environments (MPE), and the Melting Pot Harvest scenario demonstrate that $Φ$-AC learns efficient and stable coordination strategies across diverse mixed-motive settings while maintaining high collective return and competitive fairness.

Effective demand response in smart microgrids requires prosumers to cooperate voluntarily under strategic self-interest, a coordination problem structurally equivalent to a repeated Prisoner's Dilemma on a social network. This paper presents a multi-agent simulation in which a Large Language Model (LLM) Influence Compiler issues structured demand-response directives to a population of heterogeneous prosumer agents, each governed by a hybrid decision architecture combining game-theoretic base probability (derived from payoff history, neighbour imitation, and exploitation memory) with LLM narrative evaluation of incoming coordination signals. The hybrid architecture resolves a key methodological challenge: LLMs aligned via Reinforcement Learning from Human Feedback (RLHF) exhibit strong cooperation bias when used as direct decision-makers, producing flat dynamics regardless of grid conditions. By separating strategic reasoning from grounded narrative evaluation, the model generates realistic prosumer behaviour across six personality archetypes, with baseline cooperation near 50% and clear differentiation under influence. Compiled structured directives achieve 33.3% demand-curtailment cooperation versus 27.0% for unstructured messaging and 28.0% for a no-intervention baseline ($Δ_\mathrm{comp} = +0.063$), with the advantage preserved across both grounded and idealized agent substrates ($Δ= +0.083$) and across all resistance levels ($R = 0.1$ to $0.7$). Hub-targeted dissemination via high-centrality network nodes outperforms peripheral or random targeting, confirming that grid topology provides mechanistic amplification independent of message content. These results suggest that structured LLM compilation, grounded agent reasoning, and network-aware targeting are complementary design principles for scalable, interpretable demand-response coordination in smart-city energy systems.

Spiking Neural Networks (SNNs) coupled with neuromorphic hardware offer energy-efficient solutions for humanoid robot control. However, existing SNN-based motor control systems address bipedal locomotion and arm control in isolation, leaving integrated control of both unaddressed. We present a spiking architecture that coordinates force-based arm control and bipedal locomotion in a simulated humanoid, using the Neural Engineering Framework (NEF) and Semantic Pointer Architecture (SPA). High-level action selection between locomotor and arm control is mediated by a biologically grounded spiking basal ganglia model. We validate the system through co-simulation of Nengo, for the neural control, and Isaac Sim, demonstrating successful target reaching, continuous digit drawing, path-following locomotion, and finally, switching between walking and arm control via basal ganglia disinhibition. To our knowledge, this is the first integrated spiking controller to combine bipedal locomotion and arm control on a full-scale humanoid platform. The full spike-based implementation enables future deployment on low-power neuromorphic hardware.

The driving behavior of autonomous vehicles is typically governed by the cost function of their motion planner, which encodes objectives such as speed tracking, smoothness, lane keeping, and collision avoidance. However, tuning the parameters that shape this cost function is a challenging task that requires technical expertise, limiting the vehicle's ability to adapt to evolving traffic scenarios or end-user preferences. This work presents a language-driven framework for adaptive cost design in autonomous driving. A Large Language Model (LLM) interprets structured scenario descriptions and natural language user queries to generate the parameters applied to a risk-aware Model Predictive Path Integral (MPPI) controller. The system incorporates a human-in-the-loop validation stage in which the proposed behavioral changes are described in non-technical language and confirmed prior to deployment. Users may additionally provide feedback either before or after deployment, enabling iterative refinement of the vehicle's motion behavior. The framework is evaluated across multiple queries in realistic driving scenarios to assess its effectiveness. Simulation results demonstrate that the method successfully induces behavioral changes that align with the intended requirements in an intuitive manner, thereby bridging the gap between intelligent vehicle control systems and end users.

Lifelong embodied navigation in dynamic environments requires robots to form persistent scene understanding from fragmentary observations, which remains difficult for existing methods that rely on explicit maps or scene graphs and struggle to generalize beyond structured settings. We propose AllDayNav, a lifelong self-learning navigation framework that implicitly encodes scene dynamics into the billion-scale parameters of a large model via reinforcement learning, powered by a self-evolving multimodal memory that maintains and updates visual keyframes, semantic descriptions, and temporal context while autonomously generating open-vocabulary instructions, image goals, and structured rewards. Experiments in both synthetic and real-world environments across cross-room, cross-episode, and cross-task scenarios show that AllDayNav achieves success rates approaching $100\%$ and consistently surpasses strong map-based, VLM, and RL baselines in path efficiency and robustness, demonstrating implicit, memory-driven reinforcement learning as a scalable alternative to explicit mapping for reliable lifelong navigation.

Multi-terminal DC (MTDC) transmission systems based on modular multilevel converters (MMCs) are a key component of the envisioned future energy sector, where sustainability and efficiency are increasingly prioritized. To ensure their reliable operation, MMC currents must be regulated safely and rapidly under a wide range of uncertain operating conditions. Consequently, the design of current controllers faces a fundamental challenge: achieving fast transient response while maintaining robustness against uncertainties. This paper addresses this challenge by proposing a linear matrix inequality (LMI)-based design framework that leverages Lyapunov stability conditions to synthesize a less conservative static state-feedback controller. The proposed design method explicitly accounts for system constraints, including input saturation and overcurrent limits. The proposed method effectiveness is assessed on the CIGRE MT-HVDC benchmark, simulated in RTDS, and compared with existing methods.

The recent trend in scaling models for robot learning has resulted in impressive policies that can perform various manipulation tasks and generalize to novel scenarios. However, these policies continue to struggle with following instructions, likely due to the limited linguistic and action sequence diversity in existing robotics datasets. This paper introduces Task Robustness via Re-Labelling Vision-Action Robot Data (TREAD), a scalable framework that leverages large Vision-Language Models (VLMs) to augment existing robotics datasets without additional data collection, harnessing the transferable knowledge embedded in these models. Our approach leverages a pretrained VLM through three stages: generating semantic sub-tasks from original instruction labels and initial scenes, segmenting demonstration videos conditioned on these sub-tasks, and producing diverse instructions that incorporate object properties, effectively decomposing longer demonstrations into grounded language-action pairs. We further enhance robustness by augmenting the data with linguistically diverse versions of the text goals. Evaluations on LIBERO demonstrate that policies trained on our augmented datasets exhibit improved performance on novel, unseen tasks and goals. Our results show that TREAD enhances both planning generalization through trajectory decomposition and language-conditioned policy generalization through increased linguistic diversity.

Monocular local navigation is attractive for lightweight robots, but existing vision-based policies often couple perception to a specific body, camera height, and footprint, making transfer from wheeled bases to legged platforms dependent on retraining or active depth hardware. This paper introduces AgniNav, a configuration-driven local navigation framework that standardizes cross-embodiment transfer at the collision-envelope level. Each robot is specified by a measurable four-parameter safety envelope: collision-relevant height, front length, rear length, and half width. The height parameter conditions an image-to-scan network to predict a one-dimensional, collision-relevant pseudo-laserscan from a monocular color image, while the remaining footprint parameters configure a dimension-aware local planner for collision checking. Training uses height-conditioned column-minimum scan labels generated from paired color-depth data, allowing the same image to supervise different safety envelopes without collecting robot-specific data. To the best of our knowledge, AgniNav is the first monocular local-navigation framework that jointly conditions perception and planning on a shared collision-envelope configuration for zero-retraining deployment across wheeled, quadruped, and humanoid platforms. Real-robot experiments on a Turtlebot2, Unitree Go2, and Accelerated Evolution K1 achieve 39/40, 18/20, and 18/20 successes with 0/40, 1/20, and 2/20 collisions, respectively, while running at 30 Hz on Jetson Orin.

Robotic manipulation has been widely applied in industrial scenarios. Compared with single-arm manipulation, bimanual manipulation is equipped with multiple cameras to capture information from different viewpoints. However, existing multi-view policies encode each view independently or fuse view features shallowly, resulting in limited sharing semantic perception and unreliable spatial awareness. In this paper, we propose \textbf{MV-Actor}, a multi-view perception framework that builds a unified semantic-spatial representation for bimanual manipulation. First, MV-Actor performs Multi-view Semantic Interaction to share semantic perception across views. Then it uses Semantic-Spatial Token Interaction to ground visual semantics with feed-forward reconstruction model features and acquire reliable spatial awareness. Finally, a Guided Metric Depth Repair module refines degraded sensor depth to provide more reliable metric anchors under consumer-grade depth noise. In simulation experiments conducted on the PerAct2 bimanual benchmark, MV-Actor achieves a state-of-the-art average success rate of 87.8\%. In real-world evaluations with more frequent viewpoint changes and unstable consumer-grade depth, MV-Actor outperforms both RGB and RGB-D baselines, further demonstrating the benefit of sharing semantic perception and reliable spatial awareness for bimanual manipulation.

Thermo-optic phase shifters are key building blocks in Silicon and Silicon Nitride-based reconfigurable photonic integrated circuits. They enable manipulating the phase of an optical signal by means of electrically-driven heating of an optical waveguide. Conventional fabrication schemes typically require dedicated lithographic steps to separately define the resistive heaters, the current transmission lines, and the electrical contact pads. This increases the process complexity and slows the standard complementary metal-oxide-semiconductor (CMOS) fabrication flows. In this work, we present a single-step lithographic process for the realization of Titanium Nitride thermo-optic phase shifters and Aluminum interconnects integrated on a Silicon Nitride photonic platform. A detailed electro-optical characterization, performed on two platforms operating at 810 nm and 1550 nm, revealed $π$-shift powers of 92 $\pm$ 2 mW and 120 $\pm$ 10 mW, respectively. Alongside, modulation bandwidths of 8.5 $\pm$ 0.3 kHz and 3.83 $\pm$ 0.03 kHz were extracted from combined frequency- and time-domain analyses. Our results demonstrate that the proposed single-step lithographic metal definition process represents a robust, viable and cost-efficient route towards CMOS-compatible reconfigurable Silicon Nitride photonics.

Learning-based visual navigation for legged robots typically relies on continuous goal updates from hierarchical state estimation to provide a persistent directional reference. This reliance incurs additional sensory and computational overhead and deviates from fully end-to-end mobile autonomy. Furthermore, under partial observability, policies are prone to learn myopic behaviors, easily becoming trapped in dead ends and complex structural layouts. To address these limitations, we investigate a goal-initialized navigation setting, where the target is provided only once at the beginning of an episode, requiring the robot to operate based on intrinsic spatial memory without subsequent goal updates from external modules. In this work, we propose GUIDE, a fully end-to-end reinforcement learning framework designed to cultivate internal directional awareness. Specifically, GUIDE incorporates a spatial anchor predictor that leverages multi-frequency proprioceptive history to extract egomotion representations, thereby maintaining a persistent long-horizon spatial context for navigation. Concurrently, it utilizes raw depth streams to perceive local environmental geometry. We evaluate the proposed framework across both simulation and real-world scenarios on a quadruped robot. Experiments show that GUIDE learns reliable egomotion and directional awareness, enabling a fully end-to-end deployed policy to safely navigate through dense clutter and structured mazes without subsequent goal guidance or prior maps.

Real-world robotic manipulation tasks often involve forceful interactions with the environment, such as using tools of varying weights, transporting objects with different masses, and performing contact-rich tasks like table wiping. Previous learning-based approaches typically employ imitation learning policies that output target end-effector poses tracked by low-level impedance controllers. In these systems, forceful interactions are either implicitly realized through steady-state tracking errors or explicitly commanded using wrist force/torque or tactile sensors. However, implicit approaches generalize poorly across object weights, while explicit approaches require specialized hardware and increase system complexity. In this work, we propose IMPACT, a framework that decouples these forceful tasks into task-planning and internal-model-based predictive control. Extensive simulation and real-world experiments demonstrate that the proposed framework achieves higher success rates and improved generalization to unseen object weights, as well as better safety and energy efficiency.

Robot Operating System 2 (ROS 2), the de facto standard middleware framework for robots, runs each robot as a graph of nodes communicating over the Data Distribution Service (DDS), a publish/subscribe substrate. Observing this inter-node communication in real time is essential to robot development, yet it has a price. A tool can receive data only by joining the DDS domain as a subscriber that discovery has matched to the publisher, so observing folds the tool into the system it measures and perturbs it. We define this protocol-inherent perturbation as the observer's probe effect. It inflates the discovery plane, adds deserialization cost on the observer, makes the loss it reports diverge from what the subscriber actually received, and near saturation displaces the subscriber's messages. The only escape, capturing all wire traffic passively, discards ROS 2 message semantics and scales with total traffic, not what is observed. We present ros2probe, a non-intrusive observation framework that removes the probe effect. It reconstructs the full ROS 2 communication state from the domain's discovery packets at no bandwidth cost, then drives an in-kernel filter restricted to the topics the user asks for, lifting only those packets at minimal cost and observing what the real subscriber receives. Its interfaces and recordings match the standard ROS 2 tools. Across three hardware platforms (laptop, Jetson, and Raspberry Pi), two DDS implementations, and seven robot-operation workloads, ros2probe holds the discovery graph within 0.5% of an unobserved system, whereas domain-joining tools inflate discovery up to 2.6$\times$ and drop 38.5% of the subscriber's messages at saturation while ros2probe drops none. It reports loss with a recall of 1.0, cuts observer CPU and memory by up to 7$\times$ and 28$\times$, and stays practical on the embedded robots where existing tools overload the system.

Learning from human video demonstrations remains challenging due to noisy hand-object interactions, unseen objects with partial observation, and cross-embodiment discrepancy. To address these challenges, we present \textit{HOWTransfer} (\emph{H}and-\emph{O}bject \emph{O}pen-\emph{W}orld Transfer), a hand-centric framework that distills human demonstrations into contact-aware, taxonomy-informed, and diverse robotic trajectories. Instead of relying on object-specific descriptions, vision-language queries, or explicit object-state tracking, \emph{HOWTransfer} recovers temporally consistent 3D hand motion and localizes temporal contact intervals by reasoning over observed hand-object interaction cues. The localized contact onsets are then used to retarget human grasp intent into multi-modal parallel-jaw grasp hypotheses, which are propagated along the recovered wrist trajectory to generate robot-executable motions. Finally, a trajectory editing stage refines contact alignment and produces diverse executable variants from a single demonstration. Experiments across diverse manipulation tasks show that \emph{HOWTransfer} enables accurate contact localization and high-quality robot motion retargeting with $86\%$ success, which is preferred over teleoperated trajectories in a blinded preference study.

Dexterous hands are essential for fine-grained manipulation, but their hardware designs vary substantially across embodiments. Differences in kinematics, joint definitions, and degrees of freedom make it difficult to define a shared state representation compared with parallel grippers. As a result, dexterous-hand data remains fragmented and difficult to use for joint training. In this work, we propose the Unified Dexterous Hand Model (UDHM), which maps human and robot hand states into a shared 22-DoF semantic interface. Based on UDHM, we introduce UniDexTok, a retargeting-free state tokenizer that learns embodiment-conditioned discrete tokens from standardized real joint states. UniDexTok provides a unified representation for heterogeneous dexterous hands without relying on retargeting or simulation data. Compared with the recent baseline UniHM, UniDexTok reduces MPJAE from 15.63 degrees to 0.16 degrees and MPJPE from 18.51 mm to 0.18 mm, corresponding to error reductions of 98.98% and 99.03%, respectively. These results improve reconstruction from centimeter-scale to sub-millimeter accuracy. Experiments further show that data from other embodiments improves target-embodiment reconstruction accuracy, demonstrating the benefit of cross-embodiment tokenization. UniDexTok also shows strong zero-shot and few-shot reconstruction ability when new dexterous hands are introduced.

Multi-agent systems (MAS) can scale large language model reasoning at test time by decomposing complex problems into parallel subtasks. However, most existing MAS rely on centralized orchestration, where a main agent assigns work, collects outputs, and merges results. As the number of subtasks grows, this controller becomes a communication and integration bottleneck. We propose Decentralized Language Models (DeLM), a MAS framework that decentralizes coordination through parallel agents, a shared verified context, and a task queue. Agents asynchronously claim subtasks, read accumulated progress, perform local reasoning, and write back compact verified updates. The shared context acts as a common communication substrate, enabling agents to build on one another's verified progress without routing every update through a central controller. Empirically, DeLM improves both software-engineering test-time scaling and long-context reasoning. On SWE-bench Verified, DeLM achieves the best performance across Avg.@1, Pass@2, and Pass@4, with gains of up to 10.5 percentage points over the strongest baseline, while reducing cost per task by roughly 50%. On LongBench-v2 Multi-Doc QA, DeLM achieves the highest average accuracy across four frontier model families, improving over the strongest baseline by up to 5.7 percentage points. The code is available on our project website at https://yuzhenmao.github.io/DeLM/.

Rendezvous is a critical task for multi-agent systems, requiring agents to coordinate to meet at an unspecified location. However, achieving this in fluid environments presents a challenge, as it remains unclear how agents can exploit underlying fluid kinematics to facilitate convergence. In this study, we adopt a multi-agent reinforcement learning (MARL) approach to develop physics-informed rendezvous strategies in vortical flows. Compared to a naive strategy, where agents navigate toward their counterparts, MARL strategies significantly improve the rendezvous rate. MARL strategies also show transferability across varying vortex intensities, vortex scales, and swarm sizes. By breaking the symmetry of the state-action map, MARL strategy leverages a non-intuitive mechanism that prevents agents from becoming trapped in separate vortices, thereby enhancing rendezvous success. Additionally, a heuristic strategy is extracted from the learned strategy and also outperforms the naive strategy. Furthermore, a theoretical analysis demonstrates that fluid deformation impedes the rendezvous process. Large finite-time Lyapunov exponents identify where fluid effects separate adjacent agents, suggesting that targets should be planned in weak-deformation regions. Our findings reveal the important role that agent-fluid interactions play in multi-agent tasks and highlight the MARL capability to explore swarm intelligence in complex flow environments.

Robotic foundation models pre-trained on human demonstration videos have shown promise, but a significant embodiment gap remains when the resulting policies are deployed on real robots. A common remedy is to fine-tune these models on robot-specific demonstrations. However, robot data collection can be prohibitively expensive and time-consuming, which is particularly acute in dexterous manipulation, e.g., teleoperating a multi-fingered hand for even a single atomic task can take days. To address this, we introduce Dexterous Point Policy, a framework that learns dexterous manipulation policies directly from human videos and requires no robot demonstrations. Our core insight is that a unified 3D keypoint representation can bridge human and robot embodiments when used for both observations and actions. Specifically, we extract 3D keypoints of task-relevant objects and human hands from raw videos, and train an autoregressive transformer over these keypoints. We observe that at the keypoint level, specifically the wrist and fingertips, human and robot behaviors closely align, enabling direct policy transfer. On a suite of real-robot tasks spanning pick-and-place and tool use, Dexterous Point Policy attains 75.0% success, whereas a state-of-the-art VLA baseline reaches only 1.0%. Furthermore, our method generalizes strongly to unseen scenarios, including multi-object environments and novel object categories.

Zero-shot vision-and-language navigation in continuous environments (VLN-CE) has recently become feasible with large vision-language models (VLMs). However, existing methods typically rely on learned waypoint predictors to propose navigable actions. This severely limits the model's action space and fails to leverage depth inputs effectively. Moreover, memory is commonly handled by accumulating long textual or visual histories with substantial irrelevant context, or by retrieving cross-episode experiences, which weakens the zero-shot setting. In this paper, we rethink zero-shot VLN-CE as an agentic interface between the VLM and the environment, and present AgenticNav, a lightweight navigation harness that exposes action, depth, and memory as callable tools. Instead of choosing from predicted waypoints, the action tool allows the VLM to directly select a target pixel in RGB observations, converting it into executable motion. Depth is exposed through an on-demand pixel-depth tool, enabling the VLM to request precise metric distances only where they matter. For memory, AgenticNav provides a compact map image summarizing the historical trajectory, paired with a recall tool that allows the VLM to selectively revisit past visual observations without overwhelming the prompt context. On the R2R-CE benchmark, AgenticNav establishes new state-of-the-art (SOTA) performance among zero-shot methods given the same VLM backbone. Real-world validation further highlights its zero-shot generalization compared to prior methods. Ablations show that our action tool design outperforms traditional waypoint predictors, and that depth tool and agentic memory further contribute to navigation performance.

Vision-language-action (VLA) models have shown strong promise for robotic manipulation, but their reliability at test time remains limited by one-shot action prediction, where even small action errors can cause grasp failure, collision, or incorrect task progression. A natural alternative is to equip VLA systems with test-time verification, allowing multiple candidate actions to be proposed and evaluated before execution. However, reliable action verification is challenging because it requires not only distinguishing subtle geometric differences between candidate actions, but also assessing whether an action makes meaningful progress toward the task goal. We present VeriSpace, a 3D-aware action verifier for test-time action selection in VLA systems. VeriSpace evaluates candidate actions through two key components: Dual-Path 3D-Injected Scene Encoding, which constructs a scene representation that jointly preserves visual semantics and explicit 3D geometry, and Spatially-Grounded Action Reasoning, which evaluates each action by reasoning over task-relevant spatial relations, geometric validity, and expected goal progress. Together, these components enable more reliable discrimination between subtle yet outcome-critical action candidates while remaining fully compatible with existing VLA policies. Experiments on public benchmarks and real-world robotic manipulation tasks show that VeriSpace consistently improves decision reliability over both underlying VLA policies and prior verification-based methods, yielding substantial gains in both in-distribution and out-of-distribution settings.

Skill documents, structured natural-language instructions that guide Large Language Model (LLM) agents, are critical to modern agent frameworks, yet LLMs struggle to write skills that actually work. On SkillsBench, human-authored skills improve pass rates by 16.2 percentage points, while LLM-authored skills provide no measurable gain. We introduce SkillAxe, a fully unsupervised framework that enables LLMs to iteratively diagnose and refine their own skills. SkillAxe decomposes skill quality into four interpretable dimensions (quality impact, trigger precision, instruction compliance with fault attribution, and solution-path coverage), producing structured improvement briefs that require no ground-truth labels, test suites, or environment rewards. On SkillsBench, SkillAxe improves pass rates by 28\% relative over unimproved LLM skills and closes 47--67\% of the gap to human-authored skills. We validate the approach as a continuous improvement engine in the wild on SpreadsheetBench, where a SkillAxe-built skill library learns from past agent trajectories and raises pass rate from 16.0\% to 52.0\% using only 22 skills.

Deploying Vision-Language-Action (VLA) models in real robotic systems requires robustness not only to semantic and perceptual variations, but also to embodiment-side faults that change how actions are physically realized. Real robots can experience joint-level changes caused by actuator degradation, hardware faults, safety limits, collision damage, or wear-induced friction. These faults are critical because they alter the action-to-motion interface of a policy, disrupting the learned closed-loop relationship between commanded actions, realized motion, and subsequent observations. In this work, we study realistic joint-level physical faults and show that VLA models are vulnerable when predicted actions are executed through a perturbed robot body. Our analysis reveals joint-dependent effects, with heterogeneous degradation in task success across affected joints. We also show that performance drops cannot be attributed solely to physical infeasibility, since feasible faults such as increased joint friction can still substantially reduce success rates and induce closed-loop execution mismatch. Motivated by these findings, we propose Joint-level Physical-fault Aware Residual Calibrator (J-PARC), a lightweight residual calibration framework built on top of a frozen VLA policy. J-PARC infers a latent joint-fault regime from recent joint dynamics and conditions a shared residual calibrator on this regime, enabling adaptive action correction across faulty joints. Experiments show that J-PARC improves robustness under joint-level faults while preserving fault-free environment performance.

Safe social navigation requires robots to distinguish people from ordinary obstacles and to react before danger becomes imminent. We show that pretrained Vision-Language-Action (VLA) models already encode pedestrian-object distinctions and future collision signals in their internal representations, but behavior cloning fails to translate these signals into socially appropriate actions. To address this mismatch, we propose SALSA, a two-stage annotation-free post-training framework: (1) social behavioral alignment bridges intermediate-layer social features to the action head and trains on counterfactual human-object scene pairs to break visual saliency shortcuts; (2) temporal safety alignment provides automatically generated future-risk supervision to enable anticipatory collision avoidance. On SCAND and real-world deployment, SALSA reduces near-collisions by 86.4% and improves social counterfactual accuracy from 53% to 93%, demonstrating that safer social navigation can be achieved by teaching VLA policies to act on representations they already possess. These results show that pretrained VLA policies can be adapted for safer social navigation by better aligning their latent representations with action generation.

Multi-agent debate frameworks have been shown to improve large language model performance in convergent tasks, but they are currently optimized in a way that heavily favors final output accuracy rather than stability of the process. During long-horizon exchanges reactive systems under sustained perturbations often experience logic degradation, argument repetition, and role drift. To structurally prevent the identity loss and maintain the process fidelity, we introduce Knowledge-Grounded Counterfactual Reasoning (KG-CFR), a dual-stage architecture that enforces a strict separation of concerns between a private, retrieval-augmented planning buffer, and a public execution layer. We assess this system in Dynamic Resource Allocation under Uncertainty (DRAU), a dedicated 1v1v1 environment, introducing diversity as distinct from standard debate settings. Over 270 completely factorial crisis simulation trajectories with stochastic environmental shocks, KG-CFR prevents judge-detected critical post-shock degradation (defined as a quality shift, $Δ\le -0.20$) in more than 95% of perturbed runs, increasing the overall argument quality from 0.694 to 0.822. Our primary contribution is the demonstration of architectural decoupling being an important factor of systemic resilience enhancement under sustained pressure without quality loss. Furthermore, we introduce custom vector metrics for discourse divergence and plan-execution alignment that provide strong, directionally consistent evidence of operational stability. Our ablation experiments suggest that the proper doctrinal grounding can be an equally important factor for argument quality, as the prospective planning. KG-CFR, according to our initial metric evaluations, reduces semantic looping, by preserving the agent's consistency with the original plan.

Humanoid robots have achieved strong locomotion capabilities, but reliable navigation on versatile terrains remains challenging because obstacle avoidance must be coordinated with dynamically feasible motion. In this work, we present GuideWalk, a unified end-to-end framework that integrates traversability-aware navigation guidance with terrain-adaptive locomotion teacher for humanoid navigation. Specifically, we introduce a navigation module that provides explicit velocity guidance, decoupling obstacle avoidance from terrain conditions to enable robust planning across diverse environments. We propose a composite teacher distillation scheme, where goal-directed commands and dynamically consistent actions are aggregated and distilled into a single policy. To further improve robustness, the distilled policy is refined with reinforcement learning and an auxiliary behavior cloning objective, which promotes exploration while preserving desirable teacher behaviors. Experiments demonstrate that GuideWalk achieves stable and effective navigation while maintaining stable humanoid locomotion.

Real-robot evaluation is essential for understanding whether learned manipulation policies can operate reliably outside curated demonstrations. This need is particularly pressing for Universal Manipulation Interface (UMI)-style policies, whose performance depends on the coupling between wrist-view observations, action representation, data collection, and physical deployment. Existing real-world benchmarks have made important progress, but they are not designed around this UMI data-to-deployment setting. We present UMI-Bench 1.0, a local-first real-robot benchmark for standardized evaluation of UMI-style manipulation policies. To the best of our knowledge, this is the first benchmark dedicated to real-world evaluation of UMI-based manipulation models. UMI-Bench aligns data collection, scene reset, policy execution, result logging, and task-factor analysis within a unified protocol. By making the full evaluation process reproducible and auditable, UMI-Bench provides a practical testbed for measuring how UMI-trained policies generalize to real physical manipulation.

Diffusion-based action generation has become a foundational component of embodied AI, but its reliance on visual conditioning leaves deployed visuomotor policies vulnerable to adversarial manipulation. Most prior attacks focus on disruption: they perturb the observation stream to reduce task success or induce erratic behavior. We study a stronger threat, Test-time Adversarial Takeover (TAKO), in which an attacker obtains a real-time steering interface over a frozen robot policy and turns it into a remotely piloted instrument. TAKO learns a small vocabulary of reusable universal patches through differentiable diffusion inference; at test time, the attacker switches among these patches in the camera stream to compose attacker-chosen trajectories. This works because the perturbation acts on the visual conditioning pathway, where the induced bias can persist through iterative generative inference. We further show that the natural targeted baseline, target-policy matching, fails because the victim policy cannot reliably supervise itself on out-of-distribution target shifts. Across four tasks (2D manipulation, simulated aerial delivery, simulated ground navigation, and physical-world ground navigation), two visual encoders (ResNet-18 and EfficientNet-B0 + Transformer), and three generative inference families (DDPM, DDIM, and flow matching), human operators achieve 100\% takeover success on attacker-defined objectives in every evaluated setting. The project page is available at https://tako-attack.github.io.

Simulation has become an essential tool for evaluating and improving vision-language-action (VLA) policies, offering scalable, reproducible, and controllable alternatives to costly real-world robot evaluation. Recent simulation benchmarks have made substantial progress on realism and diversity, yet these platforms have not been widely adopted as reliable proxies for real-world policy evaluation. In this work, we investigate this issue through the lens of sim-and-real correlation. We conduct a systematic study across multiple simulation platforms, VLA policies, tasks, and perturbation factors, measuring whether simulated evaluation preserves real-world conclusions in terms of policy ranking consistency, performance correlation, and perturbation-wise failure patterns. This analysis allows us to characterize the limitations of existing simulators and identify what kinds of simulation signals are more aligned with real-world deployment. We further examine how users should exploit simulation for policy improvement, including when simulator-based finetuning is beneficial and how the amount of post-training data affects sim-and-real alignment. Overall, our work provides a unified framework for measuring, interpreting, and improving the usefulness of simulation for VLA policies, offering guidance both for simulator designers and for practitioners who use simulation as part of the policy development pipeline.

World Action Models (WAMs) have emerged as a new powerful paradigm for embodied intelligence, learning action-relevant visual dynamics that significantly enhance generalization and robustness. However, existing WAMs still struggle with task-relevant memory in long-horizon robotic manipulation. To address this, we present HiMem-WAM, a Hierarchical Memory-Gated WAM that integrates motion-centric latent actions, high-level skill latents, and boundary-triggered memory updates. Specifically, we develop a hierarchical latent action framework that jointly learns low-level motion and high-level skill latents, providing structured temporal abstraction. Meanwhile, a boundary-aware memory gate writes compact task states at predicted skill transitions, enabling causal inference without test-time generation of future video or optical flow estimation. Evaluated on LIBERO, LIBERO-PLUS, RMBench and real-world tasks, HiMem-WAM shows that hierarchical latents improve robustness under deployment perturbations, and the memory module substantially benefits memory-dependent long-horizon manipulation.

Object navigation requires an agent to locate a target in an unknown environment through visual observations. Existing methods typically rely on open-vocabulary detectors or vision-language models (VLMs) to answer where to search, but often overlook what not to trust - which semantic cues are unreliable. Open-vocabulary perception is prone to systematic misleading evidence: false positives, outdated static priors, and repeated failed exploration due to lack of embodied verification, which contaminates mapping and decision-making. Such errors are rooted in structured object relations in real-world scenes. To address this, we propose DB-Nav, a framework that reshapes the search space via dual relational biases. It factorizes target-centric relations into an Activation Bias (propagates contextual evidence) and an Inhibition Bias (suppresses unreliable regions via perceptual confusion and action-level falsification). These biases are unified into a Relational Activation-Inhibition Exploration Graph that modulates frontier exploration values using online observations and failed accesses. Experiments on ObjectNav benchmarks show that DB-Nav significantly outperforms existing methods in success rate (SR) and Success weighted by Path Length (SPL), offering a lightweight, interpretable, and robust navigation framework without costly online VLM reasoning.

Humanoid whole-body control has made significant progress in recent years, yet existing approaches remain limited to few-skill policies with heavy reward engineering, or motion trackers that are difficult to extend to new input modalities. We argue that the key to general-purpose humanoid control is to build a scalable brain, a module capable of reasoning with diverse conditioning modalities, atop a reactive motion tracking cerebellum, mirroring the hierarchical structure of biological motor systems. Two challenges arise in realizing this vision: acquiring a vast amount of high-quality data to achieve general purpose control, and equipping the generator with the capability to condition on compositional, extensible multi-modal inputs. We present OMG, which addresses these challenges with a meticulous data curation, filtering and labeling pipeline, as well as a diffusion-based motion generation backbone that conditions on language, audio, and human reference motions. Extensive experiments validate OMG as an omni-modal whole-body controller exhibiting state-of-the-art performance, model scaling behavior and efficient adaptation to new distributions and modalities, marking a concrete step toward foundation models for humanoid robots.

Large Language Models (LLMs) in multi-turn interactions maintain evolving context rather than generating isolated responses, making them vulnerable to prompt-injection and context-poisoning attacks in which locally plausible adversarial fragments gradually distort reasoning trajectories. Existing defenses mainly filter individual outputs and often ignore context evolution across turns, leaving long-horizon reasoning exposed. Although the Model Context Protocol (MCP) standardizes context exchange and tool invocation, it functions as a passive routing layer and does not enforce contextual stability. To address these limitations, we introduce the Game-Theoretic Secure Model Context Protocol (GT-MCP), a controller-driven multi-agent method that treats context management as a closed-loop dynamical process. GT-MCP coordinates three heterogeneous LLM agents and selects outputs through a trust function that jointly evaluates causal consistency against a validated context graph, semantic agreement among agents, and distributional drift over time. When instability is detected, a rollback-based self-healing mechanism restores the validated context and prevents unsupported fragments from propagating. Empirical evaluation over 500 interaction turns under an adaptive adversarial threat model shows that contextual drift remains bounded in 99.6% of turns, with recovery required in only 0.4%. Per-turn utility remains tightly concentrated, with median = -0.19, P05 = -0.72, and P95 = 0.30; severe degradation below -1 occurs in only 0.4% of cases, and no injection attempt succeeds at the controller level. Selected outputs maintain stable win rates above 98%, and computational overhead remains predictable, with latency per token = 1.63e-3 s.

Neural combinatorial optimization (NCO) trains autoregressive policies to solve routing problems. The standard training algorithm, REINFORCE with a rollout baseline, requires maintaining and periodically updating a frozen copy of the policy for variance reduction. This baseline introduces a structural vulnerability: on harder instances, a poor baseline produces noisy gradient estimates that can destabilize training. We evaluate Group Relative Policy Optimization (GRPO), an algorithm from large language model alignment that eliminates the baseline entirely by normalizing advantages within groups of sampled trajectories. In a controlled comparison of five RL algorithms on TSP and CVRP benchmarks within the RL4CO framework, we find that: (i) GRPO avoids the training collapse observed with REINFORCE on TSP-100, where performance degrades from cost 9.8 to 52.1 immediately after the warmup phase and does not recover under extended training; (ii) at matched gradient updates, GRPO achieves solution quality within 2% of POMO, a strong AM-based multi-start baseline, while requiring no external baseline; and (iii) P3O, a pairwise preference algorithm also from the alignment literature, is competitive on TSP but shows higher variability on CVRP. These results identify GRPO as a promising baseline-free alternative for NCO, particularly in settings where baseline-dependent training becomes fragile.

Fine-tuning vision-language-action (VLA) policies for long-horizon manipulation still relies heavily on behavior cloning, which requires costly high-quality demonstrations and keeps policies near the demonstration distribution. Reward models can reduce this dependence by reweighting demonstrations and providing dense supervision for on-robot reinforcement learning (RL), but they must be dense, accurate, and general. Existing methods fall short: task-specific stage-aware models are accurate but require per-task annotations, while general vision-language-model (VLM) reward models are broadly applicable but too coarse for fine-grained long-horizon progress. We introduce RM, a multi-task stage-aware reward model that combines an action-primitive-based stage estimator with a multi-gate Mixture-of-Experts (MMoE) value head to produce dense per-step rewards across manipulation tasks. Building on RM, we further propose SPIRAL (Self-Policy Improvement via Reward-Aligned Learning), an on-policy reward-guided framework that improves VLA policies from cheap autonomous rollouts. On a 10-task benchmark, RM reduces value-estimation MSE by 80% over the strongest baselines; when used in SPIRAL, it improves task success from around 50% to near-perfect performance on Folding Shorts (58% to 100%) and Cleaning Whiteboard (50% to 90%), showing that high-quality dense rewards are key to a stable robot data flywheel. Project website: https://qianzhong-chen.github.io/sarm2.github.io/.

Language-agent "memory palace" systems anchor each memory to a world coordinate, on the intuition that geometry adds something text cannot. We make that intuition testable and report three results. First, the memory-palace default of folding spatial proximity into a linear blend beside recency and importance does not help and can hurt: in a pre-registered recall experiment the shipped blend fails its own frozen test (mean Delta-Hit@5 -0.0375, Wilcoxon p=0.306), sitting at a position-blind baseline, while a geometry-led weighting wins decisively (+0.3208, p<10^-15): geometry must lead recall when the query regime is spatial. Second, memory recall and visibility must be separated: recall is occlusion-blind by design (you correctly remember the next room behind a wall), while visibility is a perception predicate over stored geometry that the live system never computed. A one-line ray-versus-voxel digital differential analyzer (DDA), re-pointed from the gaze ray the agent already casts, supplies it: text and the live FoV cone both score 0.000 on 849 behind-wall targets while cone-plus-DDA reaches 0.982 (exact McNemar p<10^-6); coordinate recall separately resolves near-duplicate locations a cosine null cannot (1.000 vs 0.533, n=150). Third, the visibility predicate is confirmed live under a git-committed pre-registration (SPMEM-OCC-LIVE-v1: eight scripted worlds, automated oracle scoring, 96 behind-wall targets, false-visible 1.000->0.000, pooled exact McNemar p=2.5x10^-29), a run that surfaced and fixed a real relay anchor defect. We concede that occlusion-needs-geometry is near-tautological; the contribution is the measurement and isolation, separating what spatial memory must store from how it is read. These pilots power a frozen confirmatory study (SPMEM-ZERO-REAL-PREREG-v1); the full human-authored multi-world study with blind raters remains future work.

We performed a 2D finite-element electrothermal computational study of thermoelectric effects and filamentary electronic transport in Ge$_2$Sb$_2$Te$_5$ mushroom phase change memory cells during Reset and Set operations, accounting for spatial activation energy variations in amorphous Ge$_2$Sb$_2$Te$_5$ and phase-change dynamics. Reset operations with current going from the top electrode to the narrow 4 nm bottom electrode require $\sim$3x less energy and power, and $\sim$2x lower current to achieve the same Reset resistance, compared to the opposite polarity, due to thermoelectric effects. Filamentary conduction, electrical breakdown, thermal runaway, and local crystallization of amorphous Ge$_2$Sb$_2$Te$_5$ depend on current polarity and thermal boundary conditions, and determine the location, shape, and volume of the programming region, which may be significantly smaller than the semi-cylindrical mushroom region. The programming volume does not scale with contact dimensions larger than 10 nm. Larger contact areas introduce increased device-to-device and cycle-to-cycle variability due to filamentary conduction but are expected to lead to higher reliability and endurance.

We formulate a multi-agent area coverage control problem as a two-player zero-sum game between two agent groups with conflicting goals. Conventional coverage control allocates resources based on an environmental risk density field. In contrast, we generalize this metric by allowing a second group of adversarial agents to generate the spatial risk field. Coupled agent dynamics are linked through the area coverage metric, which functions as the game reward. This framework induces coupled gradient-descent-ascent controllers for the groups. Analysis of a low-dimensional case reveals a Hopf bifurcation dictated by the ratio of the groups' control gains. In the regime dominated by adversarial agents, the system is driven into a periodic chase-evasion cycle. In the regime dominated by ordinary agents, the system converges to a fixed configuration. Numerical simulations validate these theoretical insights. Finally, we characterize the Nash equilibrium conditions. Under this equilibrium, ordinary agents converge to a generalized centroidal Voronoi tessellation, whereas adversarial agents settle at their corresponding equilibrium centroids.

Submodular set functions, characterized by the diminishing-returns property, provide a unifying combinatorial framework for many subset-selection problems in decision and control. Although exact maximization is NP-hard in general, the structural properties of submodular functions enable simple greedy algorithms that achieve constant-factor approximation guarantees for monotone objectives, with randomized greedy-based variants extending such guarantees to the non-monotone case. This survey reviews the theory, algorithms, and applications of submodular optimization with a focus on systems and control. We cover the structural properties of submodular functions, including curvature and the submodularity ratio, the constraint families that arise in practice (matroids, knapsack, and $p$-systems), and the main approximation algorithms for monotone and non-monotone submodular maximization, with up-to-date approximation ratios and hardness results. We then survey applications across sensor scheduling, multi-agent coordination, robust submodular optimization, leader-follower systems, distributed submodular optimization, game theory, system theory, resource allocation, social networks, and informative path planning. The survey emphasizes practically implementable greedy-based algorithms and instance-dependent refinements via curvature and the submodularity ratio. We close with observations on canonical control-theoretic objectives: certain functionals are submodular (the log-determinant and rank of the controllability Gramian, and the log-determinant of the Kalman filter information matrix), whereas closely related objectives fail to be sub- or supermodular (the steady-state Kalman filter error covariance, and the average control energy obtained from the inverse Gramian). We also highlight the cross-cutting open directions that follow.

Reservoir computers benefit from the inherent complexity of optical phenomena, which provide rich, often nonlinear dynamics. However, training directly on the reservoir's output renders the system prone to overfitting and computationally inefficient during the training phase. In this work, we investigate strategies to mitigate overfitting and reduce computational overhead through output pruning and regularization. We compare loss-minimizing search methods (Equal Search and Branch and Bound) against an output-oriented statistical filtering approach (Variance Filter) and random pruning, highlighting advantages and disadvantages of each approach and the overall importance of informed reservoir output sampling, particularly for a shrinking latent space. We further demonstrate that enforcing readout selection across the full output spectrum improves performance, especially for non-iterative methods. Additionally, we examine L1 and L2 regularization techniques (LASSO and ridge regression), both of which significantly enhance performance on highly nonlinear tasks such as the Spiral Benchmark. While our methods are of general use, results are obtained from and discussed exemplarily for a nonlinear fiber-optical extreme learning machine. Overall, this study provides a deep analysis of the reservoirs' hidden-layer filtering mechanisms and the output-layer training, enabling optimized performance in physical reservoir computing systems.

Foundation-model agents are increasingly long-lived systems that remember users across interactions, making memorization an explicit deployment-time function rather than solely a property of model weights. Existing work addresses parametric memorization or audits fixed memory configurations, but does not characterize how memory-design choices jointly shape personalization utility, extraction risk, and deletion fidelity. We study this surface as deployment-time memorization, formulating agent memory as a privacy-utility frontier measured by Personalization Recall (PR) and Adversarial Extraction Rate (AER), and sweeping three memory-design knobs: summarization aggressiveness, retrieval breadth (k), and deletion mode. We further introduce the Forgetting Residue Score (FRS) to quantify whether deleted information remains recoverable from derived memory tiers. On LongMemEval, key-fact summarization reduces canary extraction by 76% on Gemma 3 12B and 64% on GPT-4o-mini while preserving nearly all personalization recall; critically, once content is compressed away, increasing k no longer restores leakage. The same compression, however, induces a deletion-fidelity failure: raw-only deletion leaves derived summary copies recoverable in approximately 20% of instances, and only full-pipeline purge or tombstone redaction drives worst-tier residue to zero. Together, these results establish that persistent agent memory must be evaluated as a first-class memorization mechanism -- assessed by what it helps agents recall, what it makes extractable, and what it can truly erase.

Training reinforcement learning (RL) policies from scratch is costly: it requires careful reward and environment design, extensive tuning, and substantial computation. Yet many control problems already have a functional but suboptimal policy available as a baseline. This paper proposes a method for embedding such a baseline into the RL training process, simultaneously improving training efficiency relative to from-scratch methods and producing a learning policy that outperforms the baseline. At each step, the method arbitrates between the baseline policy and a trainable learning policy, initially relying strongly on the baseline policy and then progressively transferring agency to the learning policy. By the end of training, the learning policy is a standalone neural network that operates without baseline policy support. The paper formalizes what it means for the baseline policy to be functional: under this policy, the agent reaches a goal set and remains there with high probability. The proposed arbitration mechanism is designed to exploit this property during training, yielding high goal-reaching rates right from the beginning of training. A theoretical analysis provides a formal interpretation of this behavior under stated assumptions and extends it to the final baseline-free regime, where explicit lower bounds are derived for the goal-reaching probability of the standalone learning policy. Empirical results on continuous-control benchmarks show that the proposed method achieves returns that match or exceed those of competitive approaches, while maintaining the highest goal-reaching rates throughout training among the compared methods -- including in the final stage, where the learning policy operates without any baseline support.

Multi-agent code generation offers a promising paradigm for autonomous software development by simulating the human software engineering lifecycle. However, system reliability remains hindered by LLM hallucinations and error propagation across interacting agents. While semantic entropy provides a principled way to quantify uncertainty without ground-truth answers, current methods often rely on costly LLM-driven equivalence checks. In this work, we introduce Fast Adaptive Semantic Entropy (FASE), a novel metric that approximates functional correctness based on the minimum spanning tree of structural and semantic dissimilarity graphs. Evaluations on HumanEval and BigCodeBench demonstrate that FASE outperforms state-of-the-art semantic entropy by LLM entailment, achieving a 25% average improvement in Spearman correlation and a 19% increase in ROCAUC score against Pass@1 from ground-truth test cases when using the Qwen3-Embedding-8B model. Furthermore, by eliminating costly LLM-driven equivalence evaluation, FASE incurs negligible computational overhead, requiring only approximately 0.3% of the runtime cost of traditional semantic entropy approaches. These results position FASE as a practical, cost-effective solution for optimizing uncertainty quantification in real-world multi-agent workflows.

Optical trapping and micromanipulation rely on carefully shaped electromagnetic fields to exert precise forces and torques on microscopic particles. Despite their widespread application in biology and nanotechnology, the absolute physical limits of trapping performance, specifically the maximum achievable optical force and trap stiffness, have not yet been rigorously quantified. This work establishes a general theoretical framework to determine these fundamental bounds in the dipole approximation. By relating the optical force and stiffness to a local Taylor expansion of the electromagnetic field at the particle location, we formulate the performance limit as a solution to a quadratically constrained quadratic program. To evaluate these bounds, we employ two complementary approaches. First, we utilize a complete basis of vector spherical wave functions to determine the absolute theoretical limits of optical force and stiffness permitted by Maxwell's equations in free space, revealing Pareto-optimal trade-offs between stable confinement and directional force. Second, we introduce an aperture-based formulation that restricts the incident fields to those realizable by finite planar apertures. This yields device-consistent bounds directly applicable to experimental setups which rely mostly on electromagnetic beams. The finding that optimized aperture fields can outperform standard Gaussian beams by removing the severe axial bottleneck is particularly important. By comparing these two regimes, we identify the specific spatial modes that contribute to stable trapping and quantify the performance trade-offs inherent to physical beam shaping. This dual framework provides provably optimal bounds for power-normalized optical tweezers and serves as a rigorous benchmark for evaluating realistic beam designs.

We study a Mach-Zehnder interferometer (MZI)-based optoelectronic magnonic parametric oscillator (OEMPO) incorporating a YIG-loaded magnonic branch and a tunable phase-shifter branch, enabling systematic investigation of adaptive interferometric oscillator dynamics under distributed phase perturbations. Through analysis of nondegenerate OEPO mode pairs and frequency-pulling behavior, the loop free spectral range (FSR) and effective delay time were quantitatively extracted. Despite the nominally frequency-pinned parametric operation, weak frequency pulling and OEPO mode softening were observed, revealing an additional adaptive interferometric degree of freedom introduced by the MZI architecture. By comparing local and global sampling configurations, we demonstrate that the YIG branch behaves predominantly as a local dispersive resonant subsystem governed by the complex magnonic susceptibility, whereas the phase-shifter branch primarily controls the global interferometric redistribution geometry. Nevertheless, coherent recombination and adaptive regeneration within the loop produce finite cross-coupling between the two branches, resulting in partially synchronized interferometric dynamics and branch-dependent adaptive redistribution. Quantitative complex-Lorentzian analysis further reveals substantial phase-to-amplitude conversion and distinct differences between the OEO and OEPO regimes: the phase-pinned OEPO favors strongly dispersive local YIG response, while the frequency-adaptive OEO exhibits more mixed absorptive--dispersive behavior due to spectral relaxation through frequency pulling. Broadly, the present platform establishes a versatile framework for exploring adaptive nonlinear interferometric physics, coherent phase redistribution, and branch-dependent synchronization phenomena in hybrid magnonic-photonic oscillator systems.

Realizing the vision of 6G connected robotics requires reconciling high-performance collaborative control with the rigid spectral limitations of physical wireless channels. In realistic collaborative sensing scenarios, spectral resources are quantized into finite physical resource blocks or orthogonal subcarriers, rendering simultaneous transmission by all agents infeasible. To address this, we propose Multi-Agent Semantic K-Scheduling (MASK), a control architecture designed to sustain robust, risk-aware coordination under strict instantaneous bandwidth caps. We introduce Arbiter-Assisted Semantic Information Gating (A-SIG), a lightweight coordination mechanism that enforces hard access constraints by scheduling only the top-K agents based on locally computed semantic importance scores. By aggregating these prioritized observations into a compact latent state, a self-supervised global encoder enables a distributional policy to mitigate tail risks despite data sparsity. We evaluate MASK across diverse benchmarks, demonstrating that it matches the performance of communication-unconstrained baselines even when channel access is restricted to a small fraction of the swarm size. Furthermore, the framework exhibits inherent resilience to packet erasures, validating semantic scheduling as a critical enabler for resource-constrained 6G systems.

Reliable navigation in GPS-denied environments remains a fundamental challenge in robotics, aerospace, and autonomous vehicle applications. This paper presents a Dual Quaternion-Based Unscented Kalman Filter (DQUKF) equipped with a Visual Inertial Odometry (VIO) algorithm for accurate state estimation enabling navigation in GPS denied locations. The proposed framework formulates the DQUKF in an error state manner, where the nominal pose is represented by a unit dual quaternion and the local pose error is represented by a 6-dimensional twistor parameterization used for sigma point generation, covariance propagation, and measurement correction. In parallel, the VIO algorithm tracks features across image frames, synchronizes measurements between the IMU and camera, and provides visual constraints that complement inertial propagation. Simulation results on the EuRoC MAV dataset show that the proposed DQUKF converges under high initialization uncertainty and achieves a position RMSE of 0.2584~m in the difficult flight sequence, outperforming the benchmark filters.

Time-of-flight cameras are popular in robotics for providing direct depth information while being compact, inexpensive, and robust to lighting conditions, but their low spatial resolution and depth noise are widely believed to preclude precise feedback control. In this paper, we show that an inexpensive, low-resolution time-of-flight camera provides sufficient feedback to reliably and precisely balance an inverted pendulum on a cart--a canonical benchmark for fast, unstable dynamics.

Social Learning is a decentralized decision-making paradigm in which spatially dispersed agents collect streaming observations regulated by one of a finite number of models (the hypotheses). The agents are interested in assigning probability scores (the beliefs) to the possible hypotheses. To this end, the agents exchange their beliefs according to a certain communication graph. It has been shown that, under reasonable conditions on the identifiability of the decision model and the network connectivity, each agent ultimately places all the belief mass on the true hypothesis governing the data. However, several questions remain unanswered regarding the evaluation of the social learning performance. One recently adopted performance metric is the rejection rate, i.e., the rate at which the beliefs about the erroneous hypotheses vanish. One contribution of this work is to establish that the rejection rate leads to several paradoxes, which make it unsuitable as a valid performance measure. We then focus on studying the error probability measure. For a binary Gaussian problem, we derive an analytical formula characterizing the ratio between the individual agents' probabilities and the optimal Bayesian probability. The formula shows that this ratio is expressed by the product of two terms quantifying the effect of the network connectivity and the role of the prior information. As a result, an irreducible gap emerges between the decentralized and the centralized error probabilities, which is agent-dependent and does not disappear asymptotically.

Cloud network infrastructure at hyperscale presents unique operational challenges where traditional human-driven incident response cannot keep pace with the volume, velocity, and complexity of failures. This paper presents an agentic AI architecture for autonomous incident resolution in large-scale network operations. Our system employs a multi-agent orchestration framework where specialized AI agents collaborate to detect, diagnose, and remediate network incidents without human intervention. We describe the architectural principles, including hierarchical agent decomposition, skills-based tool invocation via standardized protocols, structured knowledge encoding from operational runbooks, progressive autonomy with safety boundaries, and closed-loop verification. The architecture has been deployed in production at a major cloud provider, demonstrating that agentic AI systems can achieve autonomous resolution rates exceeding 90% for common incident categories while maintaining safety guarantees through layered authorization and rollback mechanisms. We discuss design tradeoffs, failure modes, and lessons learned from operating autonomous AI agents at scale.

Plasmon-exciton hybrids, or plexcitons, offer deeply subwavelength light-matter interactions with versatile pathways for energy redistribution. Incorporating chirality into such systems is particularly compelling, enabling spin-sensitive optical functionality that can operate on ultrafast timescales and within ultracompact volumes. Despite recent progress in chiral plexcitonic systems, how structural chirality and plasmon-exciton coupling determine chiroptical spectra and ultrafast energy flow remains elusive. Here we realize chiral plexcitons by functionalizing intrinsically chiral gold helicoid nanoparticles with molecular J-aggregates. Within a non-Hermitian framework, we trace the microscopic origin of the helicoid chiroptical response and its coupling to the excitonic transition, revealing how the helicity of light selectively addresses distinct hybrid responses. At the spatiotemporal extreme, we find that the gap-localized response not only enhances polarization-sensitive contrast but also strengthens the local hybrid interaction, leading to accelerated ultrafast relaxation. Together, these space-, time-, and polarization-resolved measurements provide a physically grounded and experimentally benchmarked picture of chiral plexcitonic coupling, identifying chirality as a practical control parameter for selectively steering nanoscale energy pathways and dynamics.

Interior permanent magnet synchronous motor (IPMSM) design requires balancing conflicting objectives and multi-physics constraints, while modern optimization workflows face three bottlenecks: manual problem setup, high finite element analysis (FEA) cost, and unreliable surrogate-based search in sparse or out-of-distribution regions. To address these limitations, we propose an end-to-end automated IPMSM design optimization framework that integrates retrieval-augmented generation (RAG) for structured problem definition with an uncertainty-aware FEA-AI hybrid optimization pipeline. A Design agent, connected to a motor textbook through RAG, provides domain-knowledge-based options and engineering tips, and compiles an optimization card and a design-of-experiments plan for AI-model training. A Training agent automates electromagnetic FEA, records geometry-validation and solver-failure logs, analyzes failed geometries using ANOVA-based data analysis and LLM reasoning, and invokes a Design Sampling agent to redefine the design space and generate additional samples. An Optimization agent performs GA-based search with uncertainty-driven switching: low-uncertainty candidates are evaluated by AI-surrogate inference, whereas high-uncertainty and reliability-critical Pareto-front or top-K candidates are corrected by high-fidelity FEA and reused for iterative retraining. The framework converts manual, experience-dependent configuration into a reproducible workflow that balances computational cost and prediction reliability. Experimental results under a matched high-fidelity FEA budget show that the proposed hybrid approach achieves better objective performance while maintaining low and further reducible predictive uncertainty, outperforming FEA-only search, which is limited by early budget exhaustion, and AI-only search, which converges to a low-confidence optimum.

Warm starts are widely used to accelerate AC optimal power flow (ACOPF) solves, but the impact of different initialization strategies has received limited systematic study, particularly for the primal-dual interior-point methods that dominate large-scale ACOPF algorithms. This paper benchmarks initialization strategies for ACOPF solved with the interior-point solver IPOPT on 19 PGLib-OPF instances (5 to 30,000 buses), testing all 15 non-empty subsets of the primal blocks $\{P_g, Q_g, V_m, V_a\}$ under oracle conditions and three DC-seeded combinations in a practical setting. The experiments show that most partial primal-plus-dual restarts increase solve time or reduce convergence reliability. Among the oracle primal-plus-dual (O-PD) configurations, only the complete restart reliably converges on every baseline-convergent case, reaching a $47.6\%$ median solve-time speedup. Twelve of the 14 partial O-PD combinations have negative median speedups, and several fail repeatedly on larger networks. Decomposing the dual into constraint and bound multipliers shows that \emph{coverage}, not the presence of duals per se, governs robustness: the full bound-multiplier vector reaches 90.7\% convergence and a $+26.8$\% median speedup, whereas block-matched coverage (oracle multipliers on some bounds, defaults on the rest) drops to 70.4\% and $-31.1$\%. Practical DC seeding sometimes helps the AC solve, but the benefit is no longer statistically significant once the DCOPF presolve cost is included in the end-to-end comparison ($p = 0.4171$). For learned warm-start methods, the results support the following target ordering: predict the full primal vector first; if only partial coverage is possible, prioritize voltage variables; and avoid partial or inconsistent dual predictions unless the primal estimate is nearly complete.

Agent benchmarks score submissions with outcome verifiers that are typically hand-written and brittle, leaving them open to reward hacking. We audit 1,968 tasks across five terminal-agent benchmarks and find 323 (16%) hackable by frontier models given only the task description. This corrupts both leaderboard rankings and RL training signal, yet the standard response is manual and reactive. We introduce the hacker-fixer loop, a method for building exploit-resistant verifiers without per-task manual patching. The loop alternates three LLM agents: a hacker tries to pass the verifier without solving the task, a fixer patches the verifier to reject each discovered exploit, and a solver confirms the patched verifier still admits legitimate solutions. The loop iterates: each patch reshapes what the verifier rewards, surfacing the next exploit. We further add verifier access, and let patches transfer across tasks, to broaden the exploits the loop discovers. On KernelBench, the loop drives the attack success rate from 62% to 0% on a held-out corpus of publicly reported exploits. We also find that weaker agents in the loop can defend against much stronger hackers: Gemini 3 Flash's loop drives the stronger Gemini 3.1 Pro and Claude Opus 4.7's attack success rate from 76% and 61% to 0% on KernelBench, and Gemini 3.1 Pro's from 39% to 17% on Terminal Bench across 77 tasks. We release Terminal Wrench (323 hackable environments, 3,632 hack trajectories) as a snapshot of the current attack surface, our patched verifiers, the exploits the loop discovered, and our implementation as a basis for future work.

This paper proposes a design-for-test (DFT) methodology and architecture for testing and validation of silicon photonic integrated circuits. We describe the design of silicon photonic circuits and components that comprise the proposed DFT architecture. The designs are extensively simulated and validated as test-access and fault-detection circuitry. We demonstrate how the DFT approach can be deployed on photonic integrated circuits and how they can be tested for correct operation, in terms of signal power and phase. The application is demonstrated on two distinct types of designs -- an optical neural network comprising optical devices in a feed-forward topology, and on an optical logic circuit with feedback loops.

Real-world LLM applications are moving beyond single-agent workflows toward orchestrated multi-agent systems, yet current models still struggle to determine what each sub-agent needs to know. To measure this, we introduce PerspectiveGap, a benchmark for evaluating LLMs' ability to compose orchestration prompts for multi-agent systems. PerspectiveGap contains 110 scenarios, each evaluated through two distractor-mixed task formats: role-fragment assignment and free-form prompt writing. These scenarios are organized into 10 topologies, which are distilled from the authors' real-world engineering practice and framed by the Prompt Economy principle: building loop-centered orchestrations that maximize utility with minimal role and engineering overhead. In experiments with 27 commercial models from 10 companies, GPT-5.5 substantially outperforms all competitors, whereas Opus 4.7 shows a notable weakness in orchestration prompting despite its strong coding performance. Nevertheless, PerspectiveGap remains challenging: the evaluated models achieve an average combined pass rate of only 14.9\% (GPT-5.5 62.0\%) and an average overall leakage rate of 246.5\% (a per-scenario information leak-event count, not a proportion; GPT-5.5 49.1\%). These findings suggest that multi-agent orchestration prompting is a distinct and under-evaluated capability, and PerspectiveGap provides a foundation for measuring and improving it systematically.

This paper presents an adaptive model predictive control (MPC) framework for nonlinear urban air mobility (UAM) vehicles operating across the full flight envelope. The proposed approach leverages a linear parameter-varying (LPV) representation to update the predictive model online, enabling accurate capture of strongly nonlinear and time-varying dynamics associated with distributed electric propulsion (DEP) eVTOL aircraft. To systematically address the high-dimensional and coupled nature of MPC tuning, a multi-objective evolutionary optimization strategy based on NSGA-II is employed, incorporating proper normalization of states and control inputs to ensure balanced weighting and meaningful exploration of the design space. The resulting controller explicitly accounts for actuator constraints and enables reconfigurable control allocation for fault-tolerant operation. The framework is evaluated in nonlinear simulations using NASA's Generic Urban Air Mobility (GUAM) model and benchmarked against a robust servomechanism linear quadratic regulator (RSLQR). Results demonstrate that the proposed adaptive MPC achieves improved trajectory tracking and enhanced robustness under both nominal conditions and actuator degradation scenarios, including partial motor failure, while maintaining constraint satisfaction throughout all flight regimes.

Autonomous agents negotiate, purchase, deploy code, and move funds, but no neutral mechanism determines whether they met their delegated obligation, who is responsible when they did not, or which settlement action follows. This is the agentic clearing problem. Tool protocols (MCP), inter-agent communication (A2A), payment rails (x402), mandate and network agent protocols (AP2, Visa, Mastercard), and settlement-risk standards each assume that determination and none produce it. Clearing is the missing primitive. Payment is not clearing. Authorization is not clearing. LLM-as-judge evaluation is not clearing. Settlement-risk escrow is not clearing: it consumes clearing decisions. RAILS (Real-Time Agent Integrity & Ledger Settlement) is the integrity and clearing layer for agentic commerce, spanning a per-output reliability score, a published reliability record, and a clearing function that consumes them. The clearing protocol at its core closes that gap. Seven primitives (Obligation Object, Evidence Envelope, Verification Mesh, Clearing Decision, Settlement Instruction, Clearing Passport, Finality Rules), bound by a formal model of admissibility-graded verification, together yield a soundness property: no financially material settlement is supported by evidence below the obligation's admissibility floor. The property is falsifiable against the spec. We are not aware of a prior agent-commerce verification mechanism that states a property of this kind. The approaches nearest to it emit a pass, a delivery guarantee, a bare score, or an equilibrium. This paper specifies that clearing protocol.

In teleoperation, the human operator typically controls only the end-effector pose, which often leads to self-collisions of the manipulator and collisions with environmental obstacles, since joints and links are not controlled individually. A common strategy to mitigate this issue is to enhance the operator's input using optimal-control-based trajectory planning. As derivative-based solvers require differentiable constraints, existing approaches either approximate robots and obstacles with spheres, reducing geometric accuracy, or approximate derivatives, degrading convergence and increasing computation times. We address these limitations by adapting a recent formulation of differentiable collision-avoidance constraints, based on duality in convex optimization, to the teleoperation setting. The robot is approximated with capsules and the environment with polytopes. We compare the resulting trajectory planning method against state-of-the-art techniques in simulation with varying numbers of obstacles and evaluate it on a UR5e manipulator in a real-world teleoperation test. Results show that our approach achieves lower computation times while enabling more accurate obstacle modeling, leading to smoother and collision-free end-effector teleoperation.

Multirotors are widely used in applications ranging from surveillance to precision agriculture, yet conventional designs remain limited by their under-actuation. Tilt-rotor configurations overcome this limitation by enabling full actuation. This paper investigates neural-network-based control strategies for a fully actuated tilt-rotor system with four thrust-vectoring inputs. Our work is structured in two parts. First, we deliberately present a negative result by evaluating a direct input-output control approach. In this method, multilayer perceptrons (MLPs), long short-term memory (LSTM) networks, and transformer models are trained to map system states and their desired values directly to control signals. We show that this strategy fails to stabilize the system, highlighting the inherent difficulty of applying direct input-output learning to highly unstable plants. Second, as the main contribution, we propose a neural-network-enhanced sliding mode controller (SMC). The method decomposes the system dynamics into input-independent and input-dependent components, with the former learned from a small dataset using lightweight networks, thereby reducing real-time computational demands. Moreover, the proposed method can be trained using flight logs collected from low-performance controllers, and the resulting dynamic model learned from real-world data can be used in simulation. We further compare MLP- and LSTM-based implementations under model uncertainties and external disturbances, demonstrating the robustness and effectiveness of the proposed approach; in particular, the controller with the LSTM plant dynamics predictor achieves superior performance to its MLP-based counterpart while also exhibiting lower runtime.

Protecting an asset against threats is a challenging problem in an era of continuously evolving intelligent attacks. This requires cooperation between the asset and the defender to share information and jointly maneuver. To address this problem, this work proposes a cooperative guidance and control strategy for active asset protection against a maneuvering threat. This work develops a joint maneuver strategy where both the defender and the asset coordinate their time-varying speeds and courses to neutralize/capture the attacker. The control strategy is formulated around three coupled geometric and temporal objectives. The first objective is to set the line-of-sight rate between the asset and the attacker to zero, putting the attacker on a collision course and reducing their maneuvering. The second objective is to maintain the defender on the line-of-sight between the asset and the attacker. This ensures that the attacker faces the defender first before reaching the vicinity of the asset. Lastly, the defender is also guided to pursue the attacker based on the time-to-go estimates between the defender and the attacker. While keeping these objectives in mind, the control actions for the asset and the defender are jointly designed, fostering cooperation between the two. The stability of the proposed strategy is established using a Lyapunov-based approach. Numerical simulations performed show the effectiveness of the proposed cooperative strategy in ensuring the successful capture of a maneuvering threat.

We study data-driven real-time economic optimization of a multi-product chemical reactor when no reliable first-principles model is available beyond a steady-state energy balance. Instead of learning the economic objective directly as a black-box function, we use a composite formulation in which Gaussian process (GP) models predict physically meaningful outputs, including product concentrations and reactor temperature, while profit is computed analytically from these predictions together with raw-material, product, and utility prices. This preserves the structure of the economic objective, makes it parametric in changing prices without needing retraining, and allows candidate operating points to be checked against the available energy balance through a physics residual. The GPs also provide predictive uncertainty, which is exploited in a Bayesian optimization (BO) framework both for data-efficient exploration and for conservative enforcement of the reactor temperature constraint through an upper confidence bound. The acquisition function additionally penalizes large energy-balance mismatch obtained by substituting the GP-predicted outputs and candidate inputs into the available steady-state energy balance. The approach is demonstrated on a benchmark simulation of a non-isothermal multi-product reactor. Relative to a trust-region safe BO implementation, the proposed method achieves better simulated economic performance within the available iteration budget. Relative to a purely data-driven BO approach that does not use the available physics information, it avoids reactor temperature constraint violations.

I discuss some quantitative representations of Promise Theory for processes involving autonomous agents. Agent models are common in software systems, machine learning, and biology, for example, but may also apply to physics and other forms of engineering. I describe how Bayesian probability and information theoretic optimization, including Active Inference, may be incorporated with promise semantics -- as well as how Promise Theory supplements solutions, helping to avoid probability's pitfalls, which include non-local coordination, calibrating, and normalizing probabilistic computations. The role of boundary conditions in constraining allowed states and selecting decision thresholds is a form of promise, and agent alignment provides a scalable definition of intent. Autonomous agents may congeal into swarms with superagent characteristics by trying to minimize their information, despite uncertainty that works to maximize it. The use of Promise Theory involves some research challenges as well as stylistic preferences.

Multi-agent LLM systems for medical question answering often treat consensus as a reliability signal: if multiple agents agree on an answer, it is presumed trustworthy. However, answer-level consensus does not entail reasoning-level alignment. We introduce CARA (Cross-Agent Reasoning Alignment), a family of automated metrics that measure whether agents who agree on an answer also agree on the reasoning. Applying CARA to a standard debate system on two medical QA benchmarks, MedQA-USMLE and MedThink-Bench, we identify the consistency illusion: a failure mode where debate reduces detectable contradictions between agents while simultaneously decreasing the semantic similarity of their reasoning chains; agents appear to agree more but reason less consistently. To improve this misalignment, we propose the Grounded Debate Protocol (GDP), a prompt-level intervention that requires agents to commit to named medical facts and take explicit stances on other agents' claims. GDP produces large, consistent alignment improvements, with Cohen's d ranging from +1.43 to +1.99, across two datasets and two backbone models, without adding LLM calls or modifying system architecture. Our results motivate cross-agent reasoning alignment as a quantity to audit alongside accuracy in safety-critical domains.

Generating complete 3D scenes from a single image requires inferring globally consistent geometry, object relationships, and environmental context from inherently ambiguous visual evidence. Despite recent progress in joint layout-and-mesh generation, existing methods often rely on holistic or weakly decomposed pipelines that entangle many factors at once and demand extensive scene-level supervision, limiting their generalization to complex real-world environments. We propose a multi-agent orchestration framework that decomposes single-image 3D scene generation into three structured stages: scene initialization, environment construction, and multi-agent refinement. The initialization stage extracts image-derived object masks, builds object-level 3D representations, and predicts an initial spatial layout to form a coarse 3D scene. The environment-construction stage then leverages this initialization together with point-map geometry to build an environmental scaffold of supporting surfaces, room boundaries, materials, and illumination. Finally, in the refinement stage, a planner agent identifies structural and visual inconsistencies, applies simple corrections directly, and dispatches specialist agents for complex localized revisions that are reintegrated into the global scene. To provide reliable structural initialization while reducing reliance on scene-level annotations, we further introduce a geometry-aware layout predictor supervised by sparse geometric priors derived from point maps. Unlike fully supervised layout generators, the predictor can be trained from segmentation-level data and generalizes robustly to diverse real-world scenes. Extensive experiments on benchmark datasets show that our method consistently outperforms prior approaches in geometric accuracy, spatial consistency, and perceptual realism.

Adaptive beamforming is a cornerstone of array signal processing, yet its performance often collapses in the face of complex, rapidly changing interference. When interferers appear or move unpredictably, conventional estimators encounter a fundamental memory trade-off: short windows enable rapid tracking but suffer from high estimation variance, while long windows provide stable rejection but fail to adapt to shifts. This challenge is resolved by introducing the Universal Switching Beamformer (USB), which integrates competitive sequential prediction into the beamforming architecture. By employing a linear transition diagram, the USB implicitly maintains an exponentially large family of candidate covariance histories and dynamically re-weights them based on their cumulative output power. This mechanism allows the beamformer to automatically vary its effective memory length without explicit change detection or heuristic parameter tuning. A theoretical upper bound is proven on the regret relative to an omniscient oracle that selects the best piecewise-stationary covariance model in hindsight. Extensive simulations and experiments on the SwellEx-96 dataset demonstrate that the USB achieves the agility of short-window estimators and the precision of long-term integration, providing a principled solution for tracking highly non-stationary scenes.

Most evaluations of LLM agents look like exams: a discrete task, a clean environment, a score in minutes or hours. We argue that this approach is mismatched with the deployment conditions of autonomous systems, where the relevant timescale can be weeks to months, and where the dynamics that matter most, such as behavioral drift, governance in diverse environmental contexts, and cross-influence between agents from different model families, only emerge over time. We introduce Emergence World, a continuously running multi-agent simulation platform designed to make those dynamics measurable. The platform hosts populations of LLM-driven agents in a shared spatial world grounded in live external data (e.g. real-time weather, news APIs, internet access), equips each agent with 120+ specialized tools and three persistent memory systems, and lets them govern themselves through democratic mechanisms with consequential outcomes. The platform is model-agnostic at the reasoning layer and supports heterogeneous populations in which agents from different vendors share the same world. To illustrate the kinds of questions the platform makes tractable, we present a 15-day cross-vendor study with five parallel worlds powered by Claude Sonnet 4.6, Grok 4.1 Fast, Gemini 3 Flash, GPT-5-mini, and a mixed population. Identical roles and starting conditions produced radically different outcomes, ranging from stable deliberative governance to total population collapse. We release the prompts, log data and configurations to support further research on long-horizon multi-agent autonomy.

As language models are increasingly deployed as autonomous agents, they must coordinate with others over long horizons in open-ended interactive tasks. Yet existing evaluations rarely test these demands together, instead emphasising single-agent tasks, short interactions, or highly structured multi-agent settings. We introduce $alem$, a JAX-based benchmark for open-ended multi-agent coordination built on Craftax-like dynamics. Alem embeds procedurally generated coordination tasks, soft specialisation, communication, and controllable coordination difficulty into a long-horizon survival world with exploration, crafting, trading, and combat. We evaluate $13$ modern LLMs zero-shot within homogeneous teams, with trained MARL agents as reference points. Current LLM agents remain far from solving alem, averaging only ~6% normalised return, but their failures are not uniform. On the hardest coordination setting, zero-shot Gemini-3.1-Pro-High approaches MARL agents trained for one billion steps, while GPT-5.4-High achieves strong base-task reward but much lower coordination reward. This contrast shows that individual task competence does not imply coordination competence. Ablations show that communication is the largest contributor to coordination, while memory and reasoning help when used to maintain multi-step plans. Overall, our results identify coordination as a distinct bottleneck for frontier LLM agents, separate from single-agent capabilities. Alem makes this bottleneck measurable and provides a controlled testbed for developing agents that communicate, allocate roles, and execute shared plans. Code is available at https://github.com/alem-world/alem-env.