Research

• General in-context learning for embodied agents1 paper raises it · 0 address it

  Enables models to adapt to new tasks and environments without retraining, improving robustness in real-world deployment.

  • Has GPT-6-Astra Achieved Embodied Intelligence?raises
• Cost and latency in embodied control loops1 paper raises it · 0 address it

  High computational and time costs limit real-time applicability of agentic systems in dynamic environments.

  • Has GPT-6-Astra Achieved Embodied Intelligence?raises
• Physical-state verification and failure recovery1 paper raises it · 0 address it

  Agents must reliably detect and correct errors in physical interactions to avoid irreversible failures in real-world tasks.

  • Has GPT-6-Astra Achieved Embodied Intelligence?raises
• Long-horizon task preservation in embodied agents1 paper raises it · 0 address it

  Maintaining progress over extended sequences is critical for complex, multi-step tasks in robotics and navigation.

  • Has GPT-6-Astra Achieved Embodied Intelligence?raises
• Open-source stacks for generalist computer-use agents1 paper raises it · 1 addresses it

  Closed systems hinder reproducibility and systematic study of agent design for real-world computer tasks.

  • OpenGCUA: Toward a Fully Open Stack for Generalist Computer Useraises · addresses

Named by an AI from these papers' abstracts.

• Training & fine-tuning44+120.0%
• AI safety31+93.8%
• Computer vision33+83.3%
• Evals66+57.1%
• Code models14+40.0%
• Federated learning48+29.7%
• Image & video generation22+29.4%
• Robotics18+28.6%

• arXiv cs.CLOct 16

  Analyzing Repetition and Interference Effects in Memory-Constrained User Simulators

  Cara Su-Yi Leong, Nicholas Tomlin, Tal Linzen

  User simulators can help evaluate and train AI systems to interact with humans at scale. To do so, evaluation of user simulators must measure behavioral parity with people. We evaluate whether memory-constrained user simulators are sensitive to two factors known to affect human information retention: repetition of a fact and interference from competing information. We find that some simulators achieve human-like overall accuracy, and replicate the magnitude of repetition and interference effects, but none match people's accuracy on particular questions: while humans misremember facts, models forget them, leading to different error patterns. We release this data as a challenge benchmark for future work on user simulators.

  PDF
• arXiv cs.NIOct 16

  LoopHIL: Looping Hardware into LLM-assisted Agents for Wireless Embedded Systems

  Dhairya Shah, Spanddhana Sara, Savitha Viswanadh Kandala, Ambuj Varshney

  LLM agents are widely used for software development and debugging, but their adoption in embedded systems, especially for wireless tasks, remains limited. This is because embedded development requires hardware-in-the-loop testing. Agents cannot directly observe the spectrum, probe pins, or measure power, so they cannot verify wireless link functionality; currently, human engineers perform this verification. We systematically evaluate whether LLM agents can establish wireless links when provided with hardware access. We introduce LoopHIL, a framework that lets agents control a multi-device RF bench, monitor it with a spectrum analyzer, logic analyzer, and energy meter, and validate links against a verification ladder, all within code-enforced safety limits. Across four increasingly complex tasks, from generating a carrier to backscatter decoding by a commercial receiver, and using six models from two vendors, hardware-in-the-loop increases task completion rates from 54% to 79% and from 17% to 83% on the most challenging task. It also halves task time and output tokens. With debugging tools, agents decrease tag power consumption by up to 96%. However, small models often report…

  PDF
• arXiv cs.AIOct 16

  Training-Free Prompt Selection for Efficient Reinforcement Learning With Partial Rollouts

  Lei Jiang, Honghui Bao, Yue Zhou, Zhengyao Gu, Langzhou He +1

  Reinforcement learning with verifiable rewards (RLVR) has emerged as a prominent paradigm for training reasoning models, but its computational cost makes efficient data selection increasingly important. Since training examples contribute unequally to learning, selecting a small, informative subset can offer a more efficient alternative to using the entire data pool. The training-free selection method estimates prompt value with hidden states from complete greedy rollouts. However, in our datasets, $26$--$32\%$ of these rollouts fail to terminate within the generation budget. We introduce LIFT, a training-free data selector that identifies impactful prompts using only partial reasoning rollouts. At its core, LIFT uses Partial Rollout Context Lift (PRCL), a score inspired by pointwise $\mathcal{V}$-information that accumulates token-level log-likelihood ratios between a question-conditioned context and a null context with the question removed. This enables prompt scoring from a short rollout prefix rather than a complete trajectory. Across Qwen2.5-Math-1.5B and DeepSeek-R1-Distill-1.5B, prefixes retaining at least 90\% and 95\% of the full-rollout PRCL score, respectively, require…

  PDF
• arXiv cs.CLOct 16

  Language Models are Lazy Poets: Creative Mode Collapse at Scale

  Jacob A. Matthews, Sil Hamilton

  While aligned large language models (LLMs) can successfully complete many creative tasks, their outputs tend to be highly self-similar. Though solutions to this problem of "mode collapse" have been proposed, comprehensive large-scale surveys of the phenomenon in specific textual genres are still lacking. We address this gap through an extensive treatment of LLM mode collapse in poetry. Across 15 poetic forms, 248 subject matter tags, and 16 open-weight models, we generate over 1.1 million poems. To our knowledge, this constitutes the largest corpus of LLM-generated poetry to date. At million-poem scale, we find that LLM poetic vocabularies and rhyming abilities are substantial, but severely underutilized. In other words, LLMs are "lazy" poets: despite possessing large poetic inventories, LLMs over-rely on a small number of words, opening phrases, and rhymes. Further, we find that lexical homogenization differs across models and poetic forms, noting observed rhyming vocabulary is correlated with model size.

  PDF
• arXiv cs.LGOct 16

  Hidden Trajectories Don't Lie: Correcting Corruption Drift for Test-Time Adaptation

  Amirhossein Sabour, Amir Khasahmadi, Mengliu Zhao

  Test-time adaptation (TTA) adapts pretrained models to distribution shifts encountered during deployment. However, most existing methods rely on the weight representations of the final layer or signals derived from the model's output probability distribution. We study hidden trajectories: the ordered sequence of intermediate representations produced across a deep neural network. We observe that clean and corrupted trajectories become more separable as corruption severity increases. We introduce Drift-TTA, which adds zero-initialized low-rank adapters to a frozen backbone and trains them to align corrupted hidden trajectories with those from cleaner inputs. The approach supports two variants: a source-dependent version trained with clean targets, and a source-free setting that learns from corrupted images at different severities. We develop an online TTA approach that freezes the backbone and adapters while updating a scalar gate to control adaptation strength. We further cast trajectory alignment as MAP inference under a Gaussian observation model and derive a KL-based transfer bound characterizing test-time trajectory misalignment in terms of offline residual error, residual…

  PDF
• arXiv cs.CVOct 16

  Landmark-Free and Scale-Blind Multi-View Registration of Preoperative Liver Models to Intraoperative Surfaces

  Jiaming Feng, Gerardo Loza Galindo, Shahid Farid, Sharib Ali

  Augmented-reality guidance in laparoscopic liver resection requires aligning preoperative 3D liver models from CT/MRI with intraoperative surfaces. Monocular reconstruction in most cases cannot guarantee scale, albeit rigid-motion registration requires scale as an input. Our proposed framework ``SimCompose'' estimates the full similarity transformation $\Sim$, including scale, without relying on landmarks and applies the same transformation across multiple views with known relative geometry. However, reliable detection of registration failures remains challenging in surgical settings, where ground-truth alignment is generally unavailable. For this, we propose a geometry-based selector that evaluates candidate transformations without requiring ground truth. Large disagreement among these transformations is used to flag potentially unreliable registrations. Finally, we propose to use a non-rigid refinement based on a low-dimensional deformation model from Neo-Hookean simulations in the Simulation Open Framework Architecture (SOFA), without additional network training. SimCompose is trained exclusively on synthetic partial views. Its encoder processes point clouds normalised to a…

  PDF
• arXiv cs.SEOct 16

  SWE-Bench Pro V2: A Cleaner, Harder-to-Game Coding Benchmark

  Soham Dan, Jeff Da, Sanyam Bhutani, Miguel Romero Calvo, Ivan Guevara +6

  SWE-Bench Pro is a widely used benchmark of long-horizon software engineering tasks for coding agents. As agents improve, they reach task issues earlier agents never touched and interact with the evaluation environment in ways the original harness did not anticipate. We present SWE-Bench Pro V2, a revised public split built to keep the benchmark accurate as capabilities advance. Agentic audits, expert adjudication and blind re-implementation repaired or retired 480 of the 731 tasks: 23 contracted engineers spent 404 hours, each working in a language they write professionally, and every revised instruction was verified by a second engineer solving the task from the text alone. A hardened protocol restricts network access to the model endpoint and re-grades every patch in a clean container, so agents cannot retrieve upstream fixes or get credit for changes outside their patch. On the 642 V2 tasks frontier agents resolve up to 99.4% while other systems stay clearly separated, and a 51-task Hard split spreads systems from 52.9% to 98.0%. We release the benchmark and distill lessons for maintaining agent benchmarks over time.

  PDF
• arXiv cs.CVOct 16

  PREEMPT: Proactive Robustness Evaluation under Editing for Media Protection Techniques

  Qiuyu Tang, Aparna Bharati

  Proactive image protection methods aim to defend visual content against unauthorized manipulation and generative misuse by embedding protective signals into images before release. Evaluating whether these protections remain effective under image editing is therefore critical. However, existing evaluations remain largely limited to one-shot edits, providing only a static view of robustness while overlooking its evolution under realistic multi-step editing workflows. To systematically characterize robustness dynamics under sequential editing across various protection paradigms, we introduce PREEMPT, a holistic evaluation framework. It evaluates protection robustness under both one-shot and sequential editing scenarios, covering image transformations, semantic edits, homogeneous and heterogeneous edit compositions. To enable consistent analysis across protection paradigms with different underlying objectives, PREEMPT normalizes task-specific effectiveness into protection survival trajectories and summarizes long-horizon robustness through trajectory-based diagnostics. Using this framework, we reveal that degradation behaviors remain hidden under conventional one-shot evaluation.…

  PDF
• arXiv cs.SEOct 16

  Pattern-Based Precondition Inference from Method Implementations

  Shanyue Zhou, Martin P. Robillard

  Method preconditions define the conditions that must be satisfied before a method is called. Often, such conditions are not documented and must be inferred. We present Prex, a pattern-based static analysis approach that infers preconditions from explicit checks in Java source code. For each precondition, Prex reports the condition required for a valid call and the reported exception in the case of a violation. Prex also attempts code transformations to convert the condition detected within a method's implementation into an expression that can be evaluated by client code. It leverages the concept of a mutation boundary to separate precondition checks from exceptions that may occur after the method changes the program state. We evaluated Prex from 46 open source Java projects and manually reviewed independent samples of 150 reports. Among the 150 Prex reports, 149 were correct and 125 of the correct reports were executable by client code. In a systematic comparison with an alternative approach based on path expansion, Prex provided more precise results. In total, 122 of the 149 correct reports had no relevant method documentation, and 41 required nontrivial source code analysis to…

  PDF
• arXiv math.NAOct 16

  Certified neural approximations for parabolic obstacle problems

  Davood Damircheli

  Residual-trained neural surrogates carry no error guarantee, a limitation for free-boundary problems such as phase change and contact. We derive guaranteed error bounds for neural approximations of parabolic obstacle problems in first-order state--flux--multiplier form. The main result bounds the flux error of the network and the energy error of its feasible projection $\max\{v,g\}$ by the constitutive residual, the initial mismatch, the complementarity defect, and the PDE residual weighted by $c_F^2/α$, without the coercivity and norm-equivalence constants of the first-order formulation. It extends the Apushkinskaya--Repin majorant for the heat operator to general coefficients, free multipliers and a flux bound. A penetration-consistency result shows that neural triples with arbitrarily small bound exist. Outward-rounded cellwise Taylor-model enclosures turn the bound into a finite computation valid on any adaptive partition. The flux and gradient errors of the projected output are certified below $0.27$ and $0.33$ of the solution scale for a one-dimensional Stefan problem and, after adaptive refinement, below one half for all three trained networks on a two-dimensional…

  PDF
• arXiv cs.IROct 16

  Candidate-dependent residual routing trades top-of-list accuracy for recall in sequential recommendation

  Han Chen, Yingrui Li

  Candidate-dependent routing lets a sequential recommender weight a user's past behaviors differently for each item it scores. We test whether this flexibility helps by adding a small residual routing module to a self-attentive baseline, in two versions that differ only in what sets the routing weights: the user's history representation or the candidate item's embedding. Across five paired seeds on the REES46 cosmetics-shop dataset, with matched training for all three models, candidate-dependent routing leaves the primary metric, normalized discounted cumulative gain at cutoff ten (NDCG@10), essentially unchanged (mean relative change +0.003%). Secondary and exploratory analyses show that this flat average hides two effects that cancel. The candidate-dependent model places more relevant items in its top ten (recall +1.81%) but ranks the first one lower (mean reciprocal rank -1.09%), and most of its new hits land at positions six to ten. Its gains come from queries with at least one target among the user's last 50 events, while queries whose targets all lie outside that window get worse. In this form, candidate-dependent routing trades accuracy at the top of the list for recall…

  PDF
• arXiv cs.AIOct 16

  Distillation, Self-Training, and the Fragility of Learned Chemical Reasoning

  Anagha Aneesh, Sebastian Ibarraran, Frank Hu, Jason Liu, Grant Rotskoff

  Automating synthesis planning for small-molecule drug design requires chemical reasoners that are accurate as well as open to inspection and adaptation, yet frontier language models remain unreliable at predicting reaction outcomes and difficult to deploy, while small open-weight models lack the necessary domain reasoning. We show that frontier supervision can serve as a one-time bootstrap for open models on forward reaction prediction. Our models, Resonance, distill an open-weight teacher's reasoning traces and then learn from their own verified outputs, turning a small base model with near-zero accuracy into a capable reaction predictor, while a stronger student surpasses its teacher through distillation alone. Accuracy, however, does not bring robust reasoning: the small student inherits its teacher's single-pass style, self-training mainly sharpens answers it could already reach, and it readily abandons correct answers under pushback, suggesting that chemical reasoning models should be evaluated on the robustness of their reasoning as well as on the accuracy of their answers.

  PDF
• arXiv cs.SEOct 16

  Towards an Intelligent Decision Engine for Smart Contract Deployment

  Oana-Otilia Captarencu, Emanuel Onica, Andrei Arusoaie

  Layer 2 (L2) blockchain solutions emerged in recent years, predominantly in the Ethereum ecosystem, as a viable option for scaling a main blockchain network (L1). However, the selection of a L2 solution for deploying smart contracts backing blockchain applications is a decision that depends on many factors, such as costs, efficiency, popularity, and stability. We propose here a systematic methodology to help with this decision making process by using a set of relevant metrics and introducing user-defined profiles that capture different priorities among these metrics. We collect the major categories of smart contracts from a community-maintained data source, and we gather transactions for a six months period from three major L2 solutions: Arbitrum, Optimism, and Base. For each contract category and each blockchain, we compute a set of metrics and scores tailored to each user profile. By applying our methodology, we identify a clear bias in the classification, showing which L2 solutions are better suited for particular application types and user profiles.

  PDF
• arXiv cs.AIOct 16

  Agents' Last Doubt: Knowing When to Act and When to Step Back in Tool-Using Agents

  Qisai Liu, Qingyue Jiao, Bangwei Guo, Wenxi Dai, Changyi Yang +5

  Tool-using LLM agents repeatedly face decision points under uncertainty, choosing between autonomous action, additional information gathering, and human escalation. These decisions require balancing task success against interaction cost and the risk of irreversible errors. Learning to make these trade-offs is challenging because existing interaction trajectories record only the outcomes of actions taken, leaving the consequences of alternative actions unobserved. To address this gap, we construct a counterfactual dataset containing 3,115 action outcomes from 1,036 decision points in ToolSandbox. At each decision point, we execute each legal action and evaluate its downstream outcomes based on task completion, interaction cost, and irreversible errors. Using this dataset, we train SURE-Act, a controller that selects actions for a frozen backbone agent at inference time. Our experiments show that learning to rank alternative actions improves offline action-selection utility by 0.0309 over a regression-only baseline. We further show that adding SURE-Act enables end-to-end gains of up to 0.0575 in ToolSandbox's Overall Score with the Qwen3-4B agent. Moreover, SURE-Act outperforms…

  PDF
• arXiv cs.LGOct 16

  On Reparameterizing the Score Function

  Mohammad Hossein Amani, Leonardo Bocchieri, Maxime Peyrard, Clément Hongler, Robert West

  Policy gradients for autoregressive sequence models differentiate each token's log-probability while holding the conditioning history fixed. This discards information about how earlier decisions shape later conditional distributions, creating a credit-assignment bottleneck when rewards are sparse. We introduce History-Reparameterized (HR) policy gradients, which replace the frozen prefix in the score function with a differentiable reparameterized or relaxed history, allowing gradients of downstream log-probabilities to propagate backward through the generated trajectory. The resulting history-pathwise term is generally biased when added directly, so we derive Optimal HR (OHR), a centered control-variate estimator with a closed-form coefficient that minimizes the trace of the gradient variance at the population level. The variance reduction is governed by temporal structure in the policy dynamics: random or weakly structured policies provide little benefit, while policies with learned sequential dependencies yield informative history-pathwise corrections. For discrete tokens, the choice of relaxation is crucial: embedding-level relaxation yields substantial variance reduction that…

  PDF
• arXiv cs.CLOct 16

  Reasoning with Context Management and External Memory

  Man-Ling Sung, William Hartmann, Jiangying Zhou

  Reasoning models maintain information about the status of their reasoning process through in-context memory. Complex problems require larger context windows and memory requirements. In this scenario, context management without caching risks discarding useful information for reasoning. We propose RCMEM, an approach that extends context management with an external memory to maintain additional information without burdening the reasoning model with additional context. RCMEM significantly reduces computation time on problems that require repeated computation of intermediate steps, while achieving better performance on complex problems. We demonstrate our approach on a few mathematics problems using model sizes ranging from 0.4 to 25 million parameters, and GPT-2.

  PDF
• arXiv cs.LGOct 16

  Bootstrap Stochastic Interpolant Tree for Inference-time Search Under Sparse Reward

  Binglin Ji, Anindya Sarkar, Yevgeniy Vorobeychik

  Tree-based samplers offer a compelling alternative for scalable inference-time search, bypassing the per-step reward and gradient dependencies that introduce systematic bias into SMC-style frameworks. Despite this advantage, contemporary tree samplers fail to achieve broad exploration. Their reliance on learned, DDPM-like transitions inherently couples neighbor states in a trajectory, thereby restricting child-parent feature divergence in the tree sampler. Consequently, these samplers suffer from restricted coverage of the target marginal distribution, severely hindering the exploration critical for efficient search under sparse reward. To overcome these limitations, we introduce a simple yet effective Bootstrap Stochastic Interpolant Tree (BSIT) sampler that decouples child node generation from explicit parent dependencies. By expanding the degrees of freedom during child node sampling, BSIT encourages diverse trajectory rollouts and unlocks broader path-space diversity-the exact catalyst needed for efficient exploration. Crucially, we show that BSIT's improved exploration accelerates convergence to the target distribution while achieving broader distributional coverage than…

  PDF
• arXiv cs.CLOct 16

  Multi-author style detection with stylistic embeddings

  Man-Ling Sung, Chak Fai Li, Damianos Karakos, Fred Bernardin, Brian Ulicny

  We investigate the use of Large Language Models (LLMs) in multi-author style change detection. Using off-the-shelf LLMs or supervised fine-tuning solely within the target domain fails to yield effective features; for this reason, we propose cross-domain transfer learning with LLM-based embeddings fine-tuned for authorship verification. We fine-tune Mistral-7B embedding model and compare it with DeBERTa, used in the state-of-the-art (SOTA) systems in PAN-25 multi-author writing style analysis challenge. Results show that DeBERTa and Mistral excel in different user-generated text scenarios. We exploit this diversity by combining both embeddings, leading to more effective style change detection.

  PDF
• arXiv cs.CVOct 16

  Precision Cotton Disease Management: Leveraging UAV Imagery and Custom YOLO Deep Learning Framework for Enhanced Detection

  1. Anwar Iqbal*1 aiqbal. rime23smme@student. nust. edu. pk 2. Kunwar Faraz Ahmad Khan1 kunwar. faraz@smme. nust. edu. pk 3. Shahbaz Khan1 shahbaz. khan@smme. nust. edu. pk 4. Khawaja Fahad Iqbal1 fahad. iqbal@smme. nust. edu. pk 5. Sara Ali1, 2 sara. baber@smme. nust. edu. pk 6. Saqib Nazir1 saqib. nazir@smme. nust. edu. pk 7. Muhammad Azhar Javed3 muhammadazhar6231@gmail. com 1 Department of Robotics, Artificial Intelligence, SMME, National University of Science +9

  Cotton sustains the textile economies of over 80 countries with an annual yield of 113 million bales, yet 30-60% of yield losses are still attributed to viral, bacterial, and fungal diseases that remain undetected until manual field scouting. To address this critical issue, the integration of Unmanned Aerial Vehicle (UAV) imagery with deep learning offers a scalable early-warning system; however, existing models suffer from low early-stage accuracy, poor transferability, and high computational cost. In this paper, we propose Cotton-Vision, a fully convolutional single-stage detector that synergizes three lightweight architectural innovations: Ghost Bottleneck modules for a 20% parameter reduction, Efficient Channel Attention (ECA) for adaptive feature recalibration, and Cross-Stage Partial (CSP) fusion for enriched gradient flow. The network is trained on a novel dataset collected across seven cotton fields in the Punjab province of Pakistan, comprising over 2,700 high-resolution drone images. Extensive data augmentation (rotation, shear, exposure, and saturation) expands the dataset to 11,704 training samples. On a held-out test set, Cotton-Vision achieves a precision of 97.7%,…

• arXiv cs.AIOct 16

  Self-Improving Video-to-Action Models

  Hyunwoo Ryu, Eric Ming Chen, Max Wilcoxson, Vincent Sitzmann

  Video generative models are an emerging tool for embodied agents: they can serve as scalable visual planners, generating videos that depict how to complete previously unseen tasks. However, executing these plans typically requires a video-to-action or "inverse dynamics" model trained on large, human-collected video-action datasets, severely limiting scalability. We introduce Mirror, a self-supervised approach that learns video-to-action mappings through environment interactions, without prior action data or hand-crafted rewards. Mirror learns by attempting to follow a visual plan by executing its best guess of the underlying actions in the environment. Mirror then trains on the resulting video-action pairs, incrementally self-improving and following the visual planner more closely. Across Atari, OGBench, and DeepMind Control, Mirror outperforms leading learning-from-observation methods with learned video plans, and achieves near-expert performance with oracle plans. On Minecraft BASALT, it outperforms methods with access to expert actions or hand-engineered task knowledge. In Minecraft survival, it is the first method to acquire an iron pickaxe through self-improvement alone,…

  PDF