id string | title string | categories string | primary_category string | published string | url string | abstract_preview string |
|---|---|---|---|---|---|---|
2610.39070v1 | When the Same Prompt Gives Different Answers (and the Same Answer Isn't Always Right): A Taxonomy of Failure Mechanisms, Testing, and Mitigation Strategies for LLM-Assisted Evidence Synthesis Read-and-Evaluate Tasks | cs.AI | cs.PF | stat.ME | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39070v1 | LLM read-and-evaluate tasks include study screening, data extraction, risk-of-bias assessment, and critical appraisal in the evidence synthe... |
2610.39072v1 | PhysicsNeMo-Mesh: GPU-Native Meshes and Simulation Cases for Physics AI | physics.comp-ph | cs.CG | cs.LG | physics.comp-ph | 2026-10-29 | http://arxiv.org/abs/2610.39072v1 | PhysicsNeMo-Mesh is the mesh module of the open-source PhysicsNeMo library. It provides mesh processing similar to PyVista, but on the GPU,... |
2610.39073v1 | Certified Access: Test-Time Compute on Conservative Physics-Structured AI Memory | cs.AI | hep-ph | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39073v1 | Test-time compute mechanisms, such as retries, confidence-based escalation and non-local routing, are conventionally deployed as black-box a... |
2610.39074v1 | HERA: Assessing an Autonomous Agent for Quantum Device Measurement | cond-mat.mes-hall | cs.AI | cond-mat.mes-hall | 2026-10-29 | http://arxiv.org/abs/2610.39074v1 | Large language model agents have begun to carry out laboratory measurements, but their scientific conclusions are useful only if they hold u... |
2610.39076v1 | Tactus: Tactile & Pose Sensing Glove for Natural Dexterity Data Collection | cs.RO | cs.AI | cs.AR | cs.LG | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39076v1 | Collecting naturalistic human manipulation demonstrations with accurate contact force and hand pose is essential for robot dexterous manipul... |
2610.39086v1 | HALO: Human-Grounded Granular Evaluation of Proactive Agents | cs.CL | cs.CL | 2026-10-29 | http://arxiv.org/abs/2610.39086v1 | Proactive agents must decide whether, when, and how far to assist without an explicit request. Existing evaluations often rely on automated... |
2610.39090v1 | PALATE: Pretrained Autoregressive LATent Embeddings | cs.IR | cs.LG | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39090v1 | Local-commerce and food-delivery marketplaces run large ads businesses: across hundreds of thousands of stores, merchants pay to be surfaced... |
2610.39100v1 | Generating Critic-Guided Synthetic Experience in Latent World Models | cs.LG | cs.AI | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39100v1 | Learning from offline datasets is limited in scalability, as collecting new experience is often costly or unsafe in the real world. World mo... |
2610.39109v1 | PCM: A Proof-Carrying Message Protocol for Verifiable Multi-Agent Communication | cs.CE | cs.AI | cs.CR | cs.MA | cs.CE | 2026-10-29 | http://arxiv.org/abs/2610.39109v1 | When one AI agent tells another that "the tests passed" or "a person approved this refund", the receiver has only the sender's word. A2A and... |
2610.39114v1 | From Local Recovery to Unified Generation: Discovering Triple Set in Knowledge Graphs via Observation-Anchored Latent Diffusion | cs.LG | cs.AI | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39114v1 | Triple set prediction (TSP) is to recover missing relational triples in a knowledge graph by predicting the triple set instead of isolated l... |
2610.39117v1 | Uncertainty-Guided Cross-Modal Alignment for Diffusion-Based IoT Intrusion Detection | cs.LG | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39117v1 | Diffusion models evolve noisy states toward a learned data distribution, so their sampling behavior depends on the geometry of the represent... |
2610.39119v1 | CAPTURER: A Dexterous Aerial Manipulator for Dynamic and Precise Grasping across Diverse Environments | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39119v1 | As a representative form of embodied robotic interaction, aerial grasping requires close coordination between the aerial platform and manipu... |
2610.39122v1 | Appreciation Teaches Unauthorized Execution: Self-Evolution Degrades the Robustness of Personal Agents to Indirect Prompt Injection | cs.CR | cs.AI | cs.CR | 2026-10-29 | http://arxiv.org/abs/2610.39122v1 | Self-evolving personal agents act for users and learn from user feedback. However, if an agent obeys an instruction that an untrusted third... |
2610.39124v1 | Clean Sheets and Checkpoints: Evidence Complexity for Certifying Sequential Agent Risk | cs.LG | cs.AI | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39124v1 | Dangerous-capability evaluations often test an AI agent from prepared states at ordered milestones. A clean sheet records no successes. What... |
2610.39125v1 | Tracking Task Progress Through Errors and Recovery:A Benchmark and Recurrent Distributional Model | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39125v1 | Operational strategies for long-horizon robotic tasks require reliable evaluation to determine whether execution is progressing, stalling, r... |
2610.39126v1 | Actuator-Aware Observer Inputs for Cascade ADRC of Quadrotors Under Rotor Lag, Measurement Delay and Sensor Noise | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39126v1 | In sampled cascade linear active disturbance rejection control (LADRC) of quadrotors, the extended-state observer typically receives the ins... |
2610.39132v1 | DICE: Decentralized Incentive-driven Federated Learning for Efficient & Scalable Customization of Pretrained Models in Web3 | cs.CR | cs.DC | cs.LG | cs.CR | 2026-10-29 | http://arxiv.org/abs/2610.39132v1 | Customizing Pretrained Models (PMs) for downstream tasks demands task-specific data, which burgeoning Web3 applications provide abundantly.... |
2610.39147v1 | Kinematic Causal Attribution: Black-Box Explanations of Motion versus Geometry for 4D Point-Cloud Video Models | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39147v1 | 4D perception models, which classify 3D point-cloud videos, are being built into robots, assistive monitoring and AR, yet before deploying o... |
2610.39151v1 | STRATA: Reinforcement-driven Active Sampling with Tri-paradigm Adaptation for Efficient 3D Spatial Transcriptomics | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39151v1 | Spatial transcriptomics (ST) enables joint characterization of tissue morphology and gene expression. However, 3D volumetric ST acquisition... |
2610.39155v1 | REALM: Identifying Minimal Task Assets for LLM Agents | cs.CR | cs.AI | eess.SY | cs.CR | 2026-10-29 | http://arxiv.org/abs/2610.39155v1 | Large language model (LLM) agents often receive access to more files, accounts, and services than a task requires. Prompt injection turns th... |
2610.39156v1 | Learning Erasing Equal-Length Relational Pattern Languages in the Limit from Positive Data over Binary Alphabets | cs.LO | cs.FL | cs.LG | cs.LO | 2026-10-29 | http://arxiv.org/abs/2610.39156v1 | Relational pattern languages generalize Angluin's pattern languages by requiring the strings substituted for repeated occurrences of a varia... |
2610.39158v1 | VEvo-Harness: Causally-Guarded Self-Evolving Harness for Agentic Image Generation | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39158v1 | The central bottleneck in visual generation and editing has migrated from achieving pixel-level fidelity to satisfying densely constrained u... |
2610.39160v1 | CLaW: Codec-Guided Adaptive Latent Watermarking for Traceable Diffusion Image Generation | cs.CV | cs.CR | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39160v1 | With the rapid advancement of text-to-image diffusion models, increasingly realistic AI-generated content has raised serious concerns about... |
2610.39161v1 | Joint Rigid Alignment and Morphable Reconstruction from Partial 3D Face Scans | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39161v1 | Aligning a partial, real-world 3D face scan to a canonical representation requires solving two coupled problems at once: estimating the rigi... |
2610.39162v1 | A Novel Approach to Evaluating Evaluation Metrics for Multi-Output Structured Prediction | cs.LG | cs.CL | cs.CV | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39162v1 | In multi-output structured prediction tasks, while only one ground truth label may be provided in the training data, multiple equally valid... |
2610.39167v1 | EgoContact: Evaluating Physical Intelligence with Egocentric Video | cs.RO | cs.CV | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39167v1 | Egocentric video has emerged as one of the most scalable sources of data for learning generalizable manipulation priors. Yet, despite its gr... |
2610.39168v1 | Reading-Level Verification of Kanji Words in Japanese Financial Speech Synthesis | cs.SD | cs.CL | eess.AS | cs.SD | 2026-10-29 | http://arxiv.org/abs/2610.39168v1 | Few-shot voice cloning makes it cheap to produce daily Japanese financial commentary with several synthetic speakers, but kanji words with m... |
2610.39171v1 | ACAM-J-MARS: Multi-Atlas Riemannian Spectral Representations for Decoding Advanced Meditation States from fMRI | cs.LG | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39171v1 | Inferring jhana-type advanced concentration absorption meditation (ACAM-J) states from fMRI data is challenging due to the spatially distrib... |
2610.39175v1 | PRISM: Personalized Item Representations in User History | cs.IR | cs.AI | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39175v1 | Sequential recommenders typically construct history inputs from item identifiers or content features using an embedding map shared across us... |
2610.39179v1 | Edited Endpoints as Partial Goals for Image-to-Video Generation | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39179v1 | Edited endpoint images make the target of an image-to-video transition explicit and can improve transition completion, but also preserve con... |
2610.39180v1 | EmbodiExp: Compositional Experience Transfer for Embodied Agents | cs.RO | cs.AI | cs.CV | cs.LG | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39180v1 | Experience can help embodied agents turn high-level plans into grounded actions, but transferring it across tasks remains challenging: previ... |
2610.39182v1 | Reasoning or Pattern Exploitation? Mechanistic Insights from RL-Trained Language Models | cs.AI | cs.CL | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39182v1 | Large language models (LLMs) are increasingly described as acquiring "reasoning" skills after reinforcement learning with verifiable rewards... |
2610.39183v1 | Multi-level Constrained Learning for Underwater Image Enhancement | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39183v1 | Underwater image enhancement aims to correct color and recover detail degraded by wavelength-dependent attenuation, scattering, and noise. P... |
2610.39188v1 | Read More, Reason Deeper: Latent Workspaces for Long-Sequence Recommendation | cs.IR | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39188v1 | Long-sequence recommendation can improve along two axes: \emph{evidence scaling} exposes longer histories, while \emph{reasoning scaling} ad... |
2610.39193v1 | Goal-Aware Dynamic Multi-Agent Collaboration for Long-Horizon Tasks | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39193v1 | LLM-based agents are increasingly used to complete complex, long-horizon tasks without human intervention. To make progress toward the task... |
2610.39194v1 | Zero-Complement Extrapolation: Parameter Recovery and Classification in Severely Imbalanced Logistic Regression under Rank-Deficient Regimes | stat.ML | cs.LG | stat.ME | stat.ML | 2026-10-29 | http://arxiv.org/abs/2610.39194v1 | Parameter estimation and classification in binary logistic regression deteriorate severely when the number of minority class observations ($... |
2610.39195v1 | Analyzing Condensation on Patterned Surfaces via Deep Learning | cs.LG | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39195v1 | Dropwise condensation is key to power and water systems. State-of-the-art condensation surfaces are often inhomogeneous, either patterned by... |
2610.39196v1 | Pair-Dependent Judge Errors Change the Minimax Rate of Semantic Entropy | cs.IT | cs.LG | cs.IT | 2026-10-29 | http://arxiv.org/abs/2610.39196v1 | Semantic entropy, the entropy of the meanings among a language model's sampled answers, is estimated from the verdicts of a judge that decla... |
2610.39201v1 | COGS: A Diagnostic Benchmark for Attribute-Aware Object-Level 3D Gaussian Splatting Perception | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39201v1 | 3D Gaussian Splatting (3DGS) is increasingly explored as an input representation for 3D perception. Yet it remains unclear how current 3D ar... |
2610.39205v1 | Sparse Autoencoders as Implicit Diversification Regularisers for Cross-Sectional Cryptocurrency Returns | cs.LG | q-fin.PM | q-fin.ST | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39205v1 | Factor models are central to asset pricing and portfolio construction, but neural factors are often difficult to interpret, and their effect... |
2610.39212v1 | Event2Reward: Learning Robotic Rewards from Progress Relations | cs.RO | cs.LG | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39212v1 | Learning task progress from robot demonstrations provides a practical way to construct dense shaping rewards for robotic reinforcement learn... |
2610.39214v1 | Fast rigid rotational alignment of voxelized 3-D density maps in sliced Wasserstein distance | math.NA | cs.CV | math.NA | 2026-10-29 | http://arxiv.org/abs/2610.39214v1 | We develop a fast method for the rigid alignment of voxelized three-dimensional density functions over $\mathsf{SO}(3)$ using sliced optimal... |
2610.39223v1 | Where Position Matters: Profiling-Guided Layer-Wise RoPE Allocation in Transformers | cs.LG | cs.AI | stat.ML | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39223v1 | Rotary Position Embedding (RoPE) typically uses uniform coverage across Transformer layers despite heterogeneous positional requirements. Ph... |
2610.39227v1 | XLaw-JP-27B: A Lightweight Open Model Specialized for Japanese Law | cs.CL | cs.CL | 2026-10-29 | http://arxiv.org/abs/2610.39227v1 | We present XLaw-JP-27B, an open-weight model specialized for Japanese law that runs on a single 48\,GB GPU under the Apache~2.0 licence. It... |
2610.39230v1 | Beyond LLM Wrappers: Context Engineering and the Three-Engine Architecture for Reliable Financial AI | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39230v1 | Financial AI systems must support decision-bearing questions involving payments, collections, cash flow, and compliance while ensuring numer... |
2610.39231v1 | A Sequential Decision Framework for Evaluating LLM-Guided Search | cs.IR | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39231v1 | Recent systems applying retrieval-augmented generation (RAG) and "deep research" iteratively generate queries, retrieve documents, and exami... |
2610.39232v1 | Seeing and Imagining as One: Unified Video Generation and 4D Perception for Embodied World Modeling | cs.CV | cs.RO | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39232v1 | Embodied world models should not only imagine plausible futures, but also perceive the geometric, semantic, and motion structure of the evol... |
2610.39234v1 | AgentHeal: Evaluating Goal-Preserving Resource Recovery of Web Agents under Access Barriers | cs.CR | cs.CL | cs.CR | 2026-10-29 | http://arxiv.org/abs/2610.39234v1 | Large language model (LLM) agents have demonstrated strong capabilities in automating complex information-seeking and transactional tasks ac... |
2610.39237v1 | Grounding Multimodal Reasoning with Evidence-Ablated Negatives | cs.CL | cs.CL | 2026-10-29 | http://arxiv.org/abs/2610.39237v1 | Reinforcement learning with verifiable rewards (RLVR) has become a powerful tool for improving multimodal reasoning. However, standard outco... |
2610.39244v1 | CASH: A Capability-Aware Subsidy Harness for Advertiser Subsidy Allocation | cs.IR | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39244v1 | Advertising platforms subsidize advertisers through billing discounts, and Advertiser Subsidy Allocation (ASA) chooses a subsidy level for e... |
2610.39246v1 | IMRouter: Imbalance-aware Temporal Pattern Routing for LLM-based Time Series Forecasting | cs.LG | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39246v1 | Accurate time series forecasting (TSF) is essential for Web applications, such as traffic monitoring and online resource management. Recent... |
2610.39247v1 | SocialArena: Evaluating Social Intelligence in Large Language Models | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39247v1 | We introduce SocialArena, a benchmark for systematically defining and evaluating the social intelligence of large language models (LLMs). Al... |
2610.39251v1 | Where to Learn and How Far to Roll Out: Efficient Adaptive Training for GUI Agents | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39251v1 | Trajectory-level GRPO is a common approach for training multi-step GUI agents, but repeatedly rolling out complete trajectories from the tas... |
2610.39264v1 | Online Dense Reinforcement Learning with Planning-Grounding Agent for UI Navigation on Live IOS Apps | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39264v1 | While frontier vision-language models demonstrate strong capabilities in user interface (UI) navigation, their high inference costs prohibit... |
2610.39265v1 | DeCoRL: Decomposed Consistency Reinforcement Learning for Multi-Image Composition | cs.CV | cs.AI | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39265v1 | Multi-image composition (MICo) is a critical yet challenging capability for large-scale image editing models, where consistency degrades rap... |
2610.39267v1 | RAIBO-Wagon: Real-Time Navigation System for a Legged Robot Towing a Wagon | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39267v1 | Legged robots have demonstrated advanced capabilities across a wide range of real-world tasks, yet much of this progress concerns the robot... |
2610.39273v1 | PFA: Attention Can Be Progressively Focused in Long-Sequence Generative Recommendation | cs.IR | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39273v1 | Long user behavioral sequences improve the performance of generative recommendation (GR), but computing full attention over long histories d... |
2610.39275v1 | LLM-Assisted Optimization for Simulation-Based Redesign | cs.AI | cs.CE | math.OC | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39275v1 | Engineering design frequently involves redesigning a prior solution to adapt to changes in context, such as material prices or system requir... |
2610.39276v1 | MemSteer: Turning User Histories into Supervision for Agent Memory Optimization | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39276v1 | An effective agent memory system should adapt to a user's history and task requirements. Human-annotated supervision can guide this adaptati... |
2610.39277v1 | Efficient Covariance-Aware Vicinal Latent Diffusion for Continuous Conditional Image Generation | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39277v1 | Continuous Conditional Generative Modeling (CCGM) aims to synthesize high-dimensional data conditioned on continuously valued scalar labels.... |
2610.39278v1 | WorldFlow: Joint Video and 3D Motion Prediction via Large-Scale Pretraining | cs.CV | cs.AI | cs.RO | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39278v1 | Predicting future physical states stands as a central challenge in embodied intelligence. Existing video world models leave 3D dynamics impl... |
2610.39279v1 | CuBot-STT: Simultaneously Transport and Tighten for Lattice-type Modular Reconfiguration Robot | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39279v1 | Lattice-type modular reconfigurable robots rearrange standardized modules across tasks. Dedicated transport robots reduce module weight and... |
2610.39280v1 | Cross-Controller Transfer via In-Context Behavior Adaptation | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39280v1 | Generalist robot policies increasingly learn from data collected across diverse tasks and embodiments, yet typically assume a single control... |
2610.39282v1 | What Does "Now" Mean to Your Computer-Use Agent? Stale Screenshots, Page Loads, and Which Check to Run Before a Click | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39282v1 | Computer-use agents plan on a screenshot for several seconds and then click at a position on the screen; if the page has changed in the mean... |
2610.39283v1 | PriorMaskMap: Robust Online Vectorized Map Construction with Biased Priors | cs.CV | cs.RO | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39283v1 | Online vectorized map construction provides autonomous vehicles with essential, real-time semantic and geometric scene understanding. While... |
2610.39287v1 | OpenFlyV2: An onboard vision-language-action model for aerial perception and navigation | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39287v1 | Human cognition flexibly draws on a shared perceptual foundation to understand environments, attend to referenced entities, reason about spa... |
2610.39289v1 | Federated Recommender Agents | cs.IR | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39289v1 | LLM-powered recommender agents use reasoning and memory for personalization, but centralized access to user information raises privacy conce... |
2610.39292v1 | Strongly Adaptive Online Learning Achieves Uniform Discounted Regret | cs.LG | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39292v1 | Discounted regret measures the performance of an online learner by assigning exponentially decaying weights to past losses, reflecting the g... |
2610.39293v1 | Unlocking Any-Order Generation in Pretrained Autoregressive Image Models | cs.CV | cs.LG | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39293v1 | Autoregressive (AR) image generators are powerful in principle but inflexible in practice: training with a fixed raster-scan order hardcodes... |
2610.39295v1 | There Is No Free Lunch in RLVR: Every pass@1 Gain Has a pass@k Price | cs.LG | cs.AI | math.OC | stat.ML | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39295v1 | Reinforcement learning with verifiable rewards (RLVR) makes a language model solve more problems on the first try and fewer when given many... |
2610.39300v1 | Multi-Solution Inverse Design of Pixelated Patch Antennas Using an Autoregressive Transformer | cs.LG | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39300v1 | Deep-learning-based inverse design has emerged as a promising framework for pixelated patch antenna synthesis by enabling broader exploratio... |
2610.39301v1 | Youtu-Parsing-Omni: One Encoder, One Schema, All Modalities | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39301v1 | We present Youtu-Parsing-Omni, an omni-modal parser that converts documents, images, audio, and video into structured records grounded in sp... |
2610.39302v1 | Proving and Improving the Platonic Representation Hypothesis with JEPAs | cs.AI | cs.CV | stat.ML | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39302v1 | Whether models trained on different modalities converge to the same representation of the world is central to AI. The Platonic Representatio... |
2610.39305v1 | Learning to Bound: Monotonic, Calibrated Neural Upper Bounds for MPE Inference | cs.LG | cs.AI | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39305v1 | Probabilistic graphical models (PGMs) compactly encode structured joint distributions, but Most Probable Explanation (MPE), which finds the... |
2610.39306v1 | CAD-WAM: Anticipating Geometry for Text-to-CAD Generation | cs.CV | cs.GR | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39306v1 | Generating editable computer-aided design (CAD) models from text requires selecting operations that produce the intended changes to the geom... |
2610.39308v1 | Unlimited OCR Works 2 | cs.CL | cs.CV | cs.CL | 2026-10-29 | http://arxiv.org/abs/2610.39308v1 | Unlimited OCR ushers the era of long-horizon parsing, substantially mitigating the growth of working memory as document lengthens. However,... |
2610.39314v1 | Language-Grounded Force/Torque Pretraining for Extrinsic Contact Perception | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39314v1 | Fine manipulation requires robots to recognize when a grasped object makes or changes contact with its environment, yet such extrinsic conta... |
2610.39316v1 | TaxoMFM: A Taxonomy-Aware Foundation Model for the Human Gut Microbiome | cs.LG | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39316v1 | Existing gut microbiome foundation models typically use flat taxon representations, limiting their ability to exploit known taxonomic hierar... |
2610.39319v1 | ProLoop: Progress-Aware Closed-Loop Framework via Structured State Relations for Long-Horizon Robotic Task Planning | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39319v1 | Long-horizon robotic tasks are commonly decomposed into semantic subgoals to simplify planning. However, subgoal progress in LLM-based plann... |
2610.39320v1 | BEVLight: Generalizable Traffic Signal Control from UAV Bird's-Eye Views | eess.SY | cs.LG | eess.SY | 2026-10-29 | http://arxiv.org/abs/2610.39320v1 | Learning-based traffic signal control (TSC) has achieved promising performance, but most methods rely on structured traffic states that must... |
2610.39321v1 | Satisficing Regret Minimization in Bandits with Knapsacks | stat.ML | cs.LG | stat.ML | 2026-10-29 | http://arxiv.org/abs/2610.39321v1 | We study satisficing regret minimization in bandits with knapsacks (BwK), where a learner sequentially selects randomized policies, i.e., di... |
2610.39322v1 | MemoryFusion: Cross-Temporal Memory Learning for Multimodal Video Fusion | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39322v1 | Multimodal videos contain rich temporal cues that can facilitate faithful fusion. However, existing video fusion methods mainly rely on shor... |
2610.39323v1 | Class-aware Latent Attention Replay (CLAR): Class Incremental Hazard Diagnosis for Resilient AI in the Manufacturing Industrial Internet | cs.LG | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39323v1 | Artificial Intelligence (AI) systems provide critical computation services in the Manufacturing Industrial Internet (MII), and their resilie... |
2610.39324v1 | MRC-MMAgent: Reinforcement Learning for Multi-Role Collaboration in Multimodal Tool-Using Agents | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39324v1 | Multimodal tool-using agents augment vision-language models with access to external information and capabilities, enabling them to solve kno... |
2610.39327v1 | SONIC-K1: Porting a Universal Humanoid Motion Tracker to a Small Humanoid on a Single GPU | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39327v1 | We present SONIC-K1, to our knowledge the first open-weights whole-body motion tracker for the Booster K1 humanoid, obtained by adapting the... |
2610.39330v1 | Can Image Agents Complete Visual Commissions? A Benchmark and Modular Framework | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39330v1 | Visual commissions require image agents to integrate heterogeneous source materials, preserve requirements across dependent stages, and prod... |
2610.39331v1 | Telling Is Not Seeing: Textual Personas Rarely Make Vision-Language Models Answer Like the People They Describe | cs.CV | cs.AI | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39331v1 | Vision-language models (VLMs) are increasingly prompted with personas, such as "You have low vision", to stand in for people in user studies... |
2610.39339v1 | Does the Citation Matter? Quality, Grounding, and Physician-Rubric Accuracy of Google's AI Overviews for Medical Information | cs.IR | cs.CL | cs.CY | cs.HC | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39339v1 | Google's AI Overviews (AIOs) are the first result users encounter when searching the web, shifting the work of selecting and evaluating sour... |
2610.39341v1 | RetroChessBench: A Benchmark for Reconstructing and Proving Chess Histories | cs.LG | cs.AI | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39341v1 | Frontier language models such as GPT 6 Astra have contributed to research mathematics, demonstrating substantial reasoning capabilities. Yet... |
2610.39343v1 | MulClip: A table-free activation function with monotonic and folding regimes for reservoir computing on microcontrollers | cs.LG | cs.NE | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39343v1 | This letter proposes MulClip, a lightweight activation function for reservoir computing (RC) on microcontrollers (MCUs). In a ring reservoir... |
2610.39344v1 | PicturePerfect: Diffusion-based Same-identity Facial Replacement | cs.CV | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39344v1 | With the growing success of diffusion models in computer vision, we explore their potential for same-identity facial replacement in photogra... |
2610.39352v1 | Try Before You Evolve: Ranking-Scored Trials for Recursive Self-Improvement of Recommender Systems | cs.IR | cs.IR | 2026-10-29 | http://arxiv.org/abs/2610.39352v1 | Recommenders on Web platforms such as e-commerce and music streaming need their training recipes re-tuned for each dataset. Large language m... |
2610.39353v1 | CoToan: Executable Rules and a Representation Contract for Reproducible Game AI Evaluation | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39353v1 | Game AI results are comparable only when different implementations expose the same decision problem. Yet seemingly minor representation choi... |
2610.39356v1 | Semantic Mismapping in Generalized Category Discovery: Evaluation and Mitigation | cs.CV | cs.AI | cs.LG | cs.CV | 2026-10-29 | http://arxiv.org/abs/2610.39356v1 | Generalized Category Discovery (GCD) aims to perform old-category recognition and novel-category discovery in the same unlabeled set. Early... |
2610.39358v1 | Before the Next Rollout: Diagnosing Skill Utility Shifts from Policy Updates | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39358v1 | Skill-evolution methods typically analyze execution trajectories to guide agent skill refinement. However, policy updates can change the beh... |
2610.39362v1 | More Is Better? Exploring Latent Correspondence in Federated Hypergraph Learning | cs.LG | cs.DC | cs.LG | 2026-10-29 | http://arxiv.org/abs/2610.39362v1 | Federated hypergraph learning with latent cross-client hyperedge correspondence poses fundamental identifiability and computational challeng... |
2610.39367v1 | Utility-WAM: Correcting Future-Induced Action Changes | cs.RO | cs.RO | 2026-10-29 | http://arxiv.org/abs/2610.39367v1 | World Action Models (WAMs) can condition robot actions on generated visual futures, but this additional context is neither cost-free nor uni... |
2610.39372v1 | Cord: Confidence-Anchored Order Residual Decoding for Masked Diffusion Language Models | cs.CL | cs.AI | cs.CL | 2026-10-29 | http://arxiv.org/abs/2610.39372v1 | The decoding order of masked diffusion language models (dLLMs) can substantially affect generation quality, yet common decoders typically ap... |
2610.39373v1 | What Makes Medical Examination Items Difficult? Using Large Language Models to Explore Patterns Associated With Empirical Item Difficulty | cs.CL | cs.AI | cs.CL | 2026-10-29 | http://arxiv.org/abs/2610.39373v1 | Research on medical item difficulty has largely focused on prediction rather than on identifying interpretable patterns associated with empi... |
2610.39377v1 | AgentDecept: Benchmarking Deceptive Behavior in Tool-Using LLM Agents under Pressure | cs.AI | cs.AI | 2026-10-29 | http://arxiv.org/abs/2610.39377v1 | LLM agents increasingly carry out multi-turn tasks across everyday applications, yet users often see only the agent's report rather than the... |
End of preview. Expand in Data Studio
No dataset card yet
- Downloads last month
- 338