Latest papers

119K AI papers from arXiv, newest first. The most recent is from November 3.

November 3, 2026
DreamX: Video Diffusion Model is a Universal Latent Action Learner
Nov 3, 2026Jisoo Kim, Chaehyun Kim, Junsu Kim +6Latent Action Learning Action Representation Learning
KAIST · NVIDIA
From Autonomous Agents to Cognitive Symbionts: Toward AI-Augmented Human Intelligence
Nov 3, 2026Hao FeiHuman-AI Collaboration Human-Centered AI
University of Oxford
Aligning Language Models with General Preferences: An Empirical Study of Direct Nash Optimization
Nov 3, 2026Austin Tsai, Saurav Nanda, Ganapathy Parthasarathy +2LLM Alignment Preference Optimization
Synopsys, Inc., Sunnyvale, CA, USA.
PolicyIK: High-Precision Inverse Kinematics for Robot Arms with Reinforcement Learning
Nov 3, 2026Yuhan Liu, Xinqi Liu, Zixuan Dong +2Inverse Kinematics RL for Robotics
New York University · New York University Abu Dhabi
GUILD: Guided Iterative Data Collection for Imitation Learning
Nov 3, 2026Shivin Dass, Kevin Nguyen, Rohan Yelandur +2Robotic Data Collection Imitation Learning
University of Texas at Austin
Visibility-Aware Reference Generation for Benchmarking Visual 3D Reconstruction in Vegetation-Rich Environments
Nov 3, 2026Zhipeng Zhong, Matthew Coombes, Cunjia Liu3D Reconstruction 3D Reconstruction Benchmarks
Department of Aeronautical and Automotive Engineering, Loughborough University, Leicestershire, UK, LE11 3TU
GSCV++: Lightweight 3D Gaussian Splatting Sequence Compression via Standard Video Codecs
Nov 3, 2026Qi Yang, Yingzhan Xu, Kai Zhang +13DGS Compression Video Compression
Bytedance Inc., San Diego · Bytedance Inc., Beijing
Suppressing Misaligned Behaviors in Language Models, Durably
Nov 3, 2026Luxi He, Pengcheng Jiang, Jifan Zhang +2LLM Alignment LLM Unlearning
Princeton University · UIUC · Thinking Machines Lab
Reading Why: Born-Measured Attention with Exact Word Evidence on Light and Qubits
Nov 3, 2026Teerawat Tongloy, Jiraporn Pongsopa, Rangsan JomtarakFaithfulness of Language Model Explanations
School of Integrated Innovative Technology, King Mongkut’s Institute of Technology Ladkrabang, Bangkok, Thailand · Faculty of Science and Technology, Suan Dusit University, Bangkok, Thailand
OpenBrain: An Auditable Generated-Label Release for Whole-Brain MRI Parcellation
Nov 3, 2026Qizhen Lan, Yu-Chun Hsu, Yuxiang Wei +5Brain MRI Medical Image Segmentation
The University of Texas Health Science Center at Houston · Georgia Institute of Technology · University of Houston–Clear Lake +1
Agnostic Learning by Imitating
Nov 3, 2026Sílvia Casacuberta, Parikshit Gopalan, Varun Kanade +3Agnostic Learning Statistical Learning Theory
Stanford University · Apple · Oxford University +2
Adversarial Speculative Reasoning Improves Long Horizon Agents
Nov 3, 2026Hongyin Luo, Wei Fang, Lijie Yang +5Speculative Decoding Long-Horizon LLM Reasoning
Subconscious Systems Technologies, Inc. · Massachusetts Institute of Technology · Princeton University +1
A Circuit for In-Context Learning
Nov 3, 2026Andrew Gao, William GaoIn-Context Learning Circuits in Language Models
Stanford University
Weight-space programming for auditing clinical decision networks
Nov 3, 2026Humza Nusrat, Omar NusratWeight-Space Analysis
Department of Radiation Oncology, Henry Ford Health, Detroit, Michigan 48202, USA · Department of Radiology, Michigan State University, East Lansing, Michigan 48824, USA · Toronto Metropolitan University, Toronto, Ontario, Canada
MedPhysBench: A Benchmark for Evaluating Language Models on Medical Physics Calculations and Error Detection
Nov 3, 2026Omar NusratClinical Language Model Evaluation Language Model Error Detection
Department of Physics, Toronto Metropolitan University, Toronto, Ontario, Canada · Institute for Biomedical Engineering, Science and Technology (iBEST), St. Michael’s Hospital, Unity Health Toronto, Toronto, Ontario, Canada
A simple adaptive accelerated optimistic dual averaging method for unconstrained convex optimization in $p$ -norm geometry
Nov 3, 2026Bowen Li, Danqing Zhou, Chuan HeAccelerated Gradient Methods Smooth Convex Optimization
Department of Mathematics, Linköping University, Sweden
CATF: A Manager-Worker Framework for Context-Aware Time-Series Forecasting
Nov 3, 2026Yueqiao Chen, Dongyu LiuTime Series Forecasting Time Series Forecasting with Covariates
UC Davis, Davis, CA, USA
PyRUA-DataFarm: LLM Agent as a Cost-Efficient Teleoperator for Real-Ready Robot Data
Nov 3, 2026Kewei Zhang, Juncheng Ma, Guodong Lv +11Robotic Data Collection Synthetic Data Generation for Robotics
Peking University · NVIDIA
HARP: Training-Free Dual-Profile Agentic Communication for LLM-Based Recommendation
Nov 3, 2026Zhen Tao, Xun Zhou, Xuhui Chen +2Recommender Systems Large Language Model-Based Recommendation
Harbin Institute of Technology, Shenzhen · Great Bay University · Shenzhen Loop Area Institute
MDSKILL: Learning to Internalize Skill Memory through Multi-View Distillation for Efficient Dynamic Agent Self-Evolution
Nov 3, 2026Dian Jin, Jinluan Yang, Kairong Han +9Agent Skill Learning RL for LLM Agents
Zhejiang University · Kuaishou Technology
$Φ$ asmaNet: A Staged Spectral Token-Refinement Architecture for Fast Gravitational-Wave Signal-to-Noise Ratio Prediction
Nov 3, 2026Ioannis Tsioulis, Anastasios Tefas, Nikolaos StergioulasGravitational-Wave Astronomy Tabular ML
Department of Physics, Aristotle University of Thessaloniki, Greece · School of Computing, Mediterranean College (University of Derby), Athens, Greece · Department of Informatics, Aristotle University of Thessaloniki, Greece
You don't need long context
Nov 3, 2026Peter Belcak, Saurav Muralidharan, Jindong Jiang +9Long-Context Reasoning in Language Models Long-Context Language Model Inference
LoopVQ: Sub 1-bit KV Cache Quantization for Looped Transformers
Nov 3, 2026Minsoo Cheong, Sungjoo YooKV-Cache Quantization LLM Inference Acceleration
Seoul National University
PACT: Pose-Robust Actuation Co-Design for Tendon-Driven Robotic Hands
Nov 3, 2026Prashant Kumar, Rachael Granberry, Taylor Niehues +2Robotic Hand Design Tendon-Driven Actuation
Meta Reality Labs Research
Universal Hardness From Masking Channels
Nov 3, 2026Parikshit Gopalan, Manoj Prabhakaran, Pranay Tankala
Apple · IIT Bombay · Harvard
GSBR: Geometric Support Behavior Regularization for Online RL
Nov 3, 2026Charles Xu, Jacob Berg, Liyiming Ke +1Regularized RL Robotic RL
UC Berkeley · Physical Intelligence
Wonder 4: A Mixture-of-Experts Recipe for Image Quality Enhancement
Nov 3, 2026Yunan Wu, Bin Sun, Partha Acharjee +1Image Enhancement All-in-One Image Restoration
Topaz Labs Research
Personas Are Not People: Identity as State, Not Attribute
Nov 3, 2026Tamunotonye Harry, Yuanyuan FengHuman Behavior Simulation Human Behavior Modeling
University of Vermont, Department of Computer Science, Burlington, Vermont, USA
Low-Density 3D Point Cloud Classification
Nov 3, 2026Ahmed Baha Ben Jmaa, Faten ChaiebPoint Cloud Classification Point Cloud Learning
Efrei Research Lab Paris Panth´eon-Assas University, France
False Reassurance in LLM Severe-Weather Guidance: A Replayable Evaluation Against Real NWS Alerts
Nov 3, 2026Kadharmoideen FadurudeenLLM Reliability Language Model Safety Evaluation
Independent Researcher
Knowingly Wrong: Training LLMs Against Their Own Beliefs can Induce Misalignment
Nov 3, 2026Ariana M. Villegas-Suarez, Guido Freire, Sohan Venkatesh +1Emergent Misalignment in Language Models Knowledge Conflicts in Language Models
LASR Labs · Warsaw University of Technology
Anchored Rank-Adversarial Fine-Tuning Pretrained Visual Place Recognition Models
Nov 3, 2026Mostafa Kira, Bader Rasheed, Manuel MazzaraVisual Place Recognition Adversarial Robustness
Faculty of Computer Science and Engineering, Innopolis University, Innopolis, Russia.
YANAMI: A Large-Scale Multimodal Dataset and Benchmark for Multiturn Stateful Interaction
Nov 3, 2026Shengyu Wang, Tengyue Nong, Junming Liu +10Multimodal Dataset Multi-Turn Dialogue Evaluation
The Hong Kong University of Science and Technology · Independent Researcher · The University of Melbourne +7
Möbius: A Unified Motion-Language Model Bridging Understanding and Generation
Nov 3, 2026Jiazhi Shu, Chengfeng Zhao, Jinyi Wang +7Human Motion Generation Text-to-Motion Generation
The Hong Kong University of Science and Technology · Tencent Games Motus · Massachusetts Institute of Technology
Beyond Accuracy: What Language-Subspace Interventions Capture, Remove, and Improve
Nov 3, 2026Shaanil Punglia, Dadhichi Telwadkar, Vaibhav Rathore +2Causal Interventions in Language Models Cross-Lingual Reasoning in Language Models
Motilal Oswal Financial Services, India · Indian Institute of Technology Bombay, India
Conformal Prediction at Performative Equilibrium: The Price of Performativity
Nov 3, 2026Ahmed Mahrous, Roberto Di PietroConformal Prediction Performative Prediction
King Abdullah University of Science and Technology (KAUST), Thuwal 23955, Saudi Arabia
De Rham cohomological architectures for PINNs and PDEs with topological holes
Nov 3, 2026Andrew GracykNeural PDE Solvers Physics-Informed ML
Department of Mathematics, Purdue University, West Lafayette, IN 47907, United States
Route Anything: A Study of Model and Policy Routing for General-Purpose AI Systems
Nov 3, 2026Pengfei Zhou, Yixing Ma, Hexin Wang +4Adaptive Model Routing Multimodal Model Routing
National University of Singapore, HPC-AI Lab · InfRec, Cardinal AI Lab · University of California, Berkeley +1
The Geometry of Ignorance: Detecting Suppressed Knowledge in LLMs through Activation Fragmentation
Nov 3, 2026Adar Peleg, Dvir Alsheich, Tomer Ashuach +1Factual Knowledge in Language Models LLM Unlearning
Technion - Israel Institute of Technology
iGC: Antibody Optimization via Reinforcement Learning over a Learned Somatic Hypermutation Prior
Nov 3, 2026Thao Nguyen, Robert G. Alberstein, Ewa Nowara +10Protein Design RL for Generative Models
University of Illinois at Urbana-Champaign · Genentech · F. Hoffmann-La Roche Ltd.
Stabilizing On-Policy Targeted Guidance in Agentic Reinforcement Learning
Nov 3, 2026Kevin Frans, Alexander M Rush, Sergey Levine +2RL for LLM Agents Agentic RL
Cursor Research · UC Berkeley
Gaussian Partitioning for Reliable Learning of Multi-Index Models
Nov 3, 2026Ilias Diakonikolas, Giannis Iakovidis, Daniel M. Kane +1Agnostic Learning Multi-Index Models
University of Wisconsin-Madison · University of California, San Diego
JHiMA-WebRL: A Joint Hierarchical Multi-Agent Reinforcement Learning Framework for Long-Horizon Web Tasks
Nov 3, 2026Shanka Subhra Mondal, Jiayun WangRL for GUI Agents Long-Horizon Agent Tasks
Center for Advanced AI, Accenture
NovoCRF: Globally Normalized De Novo Peptide Sequencing on a Prefix-Mass Lattice
Nov 3, 2026Rui Qiao, Lei Xin, Baozhen ShanMass Spectrometry Conditional Random Fields
Bioinformatics Solutions Inc.
Going Deeper With Memories
Nov 3, 2026Afshin KhadangiLanguage Modeling Byte-Level Language Model
SnT, University of Luxembourg
Should the Action Expert Train the VLM? Action-Token Supervision and Gradient Routing in Mixture-of-Transformer VLAs
Nov 3, 2026Asher James Hancock, Boyi Li, Yuxiao Chen +6VLA Model Training Flow-Based VLA Models
NVIDIA · Princeton University · Stanford University
Reality and Perceptual Gaps in Expressive Swarm Robotics
Nov 3, 2026Ali Imran, Abdalwhab Bakheet Mohammed Abdalwhab, Rafael Gomes Braga +7Robotics Swarm Robotics
INIT Robots Laboratory, Department of Mechanical Engineering, École de technologie supérieure, Montréal, Québec, Canada . · Department of Dance, Université du Québec à Montréal, Montréal, Québec, Canada . · MIST Lab, Department of Computer and Software Engineering, Polytechnique Montréal, Montréal, Québec, Canada .
Logarithmic Regret for Alternating Mirror Descent
Nov 3, 2026Tianlong Nan, Christian KroerRegret Minimization Online Mirror Descent
Columbia University, Department of Industrial Engineering and Operations Research.
Shrinkage prior for implicit generative models
Nov 3, 2026Shreya Sinha-Roy, Christian P. Robert, Ritabrata DuttaGenerative Modeling Bayesian Inference
Department of Statistics, University of Warwick, Coventry, CV4 7AL, United Kingdom · CEREMADE, Université Paris Dauphine PSL, France
Post-stratified importance sampling
Nov 3, 2026Yuling Yao, Sergei TikhonovImportance Sampling Variance Reduction
Department of Statistics and Data Science, University of Texas at Austin
ARLoop: Fully-Open Pre-trained Looped Transformers with Attention Residuals
Nov 3, 2026Yuekun Yao, Yupei Du, Dawei Zhu +1Recurrent Transformers Attention Residual Connections
The Ohio State University · Saarland University
Diagnosis by Composition: Detecting and Localizing Unseen Diseases by Composing Learned Radiological Findings
Nov 3, 2026Kaihua Hou, Wei Liu, Yi Huang +10Medical Image Analysis Medical Image Representation Learning
University of California, Berkeley · Alibaba Group · Fudan University +4
Regret-based Characterization, Information Limits, and the Sample Complexity of Auditing Adaptive Policies
Nov 3, 2026Irene AldridgeMarkov Decision Processes Algorithmic Auditing
Post-Inference Model Chaining: A Problem Statement and a Model-based Bootstrap for Training End-to-End Digital Twins
Nov 3, 2026Lukas Madenach, Timo Bailer, Tobias EilertDigital Twins Regression Prediction Intervals
CMC Statistics BioPharma, Boehringer Ingelheim Pharma GmbH & Co. KG, Ingelheim, Germany
Learning curves in natural and artificial systems
Nov 3, 2026Dmitry A. NovikovLearning Curves Neural Scaling Laws
V. A. Trapeznikov Institute of Control Sciences of the Russian Academy of Sciences, Moscow
PLANTEUR: A Crop-Type Segmentation Benchmark for Tropical Regions
Nov 3, 2026Piper Wolters, Gabriel Tseng, Favyen Bastani +4Agricultural Remote Sensing Remote Sensing Image Segmentation
Allen Institute for AI
Fairness Metrics based on the AUC in Clinical Predictive AI
Nov 3, 2026João Matos, Ben Van Calster, Richard D. Riley +2Algorithmic Fairness Clinical Prediction
Centre for Statistics in Medicine, University of Oxford, UK · Department of Development and Regeneration, KU Leuven, Belgium · Department of Applied Health Sciences, University of Birmingham, UK +1
Fusing Earth observation embeddings and spaceborne terrain with meteorological forcing for flood peak depth prediction from high-water marks
Nov 3, 2026Wen Zhou, Shaowen Wang, Upmanu Lall +2Flood Forecasting Hydrology
Department of Geography and Geographic Information Science · CyberGIS Center for Advanced Digital and Spatial Studies · Siebel School of Computing and Data Science +2
Physics-informed traditional and generative Mori-Zwanzig-reduced models for multiscale systems
Nov 3, 2026Emily Williams, Saad Qadeer, Amanda Howard +1Reduced-Order Modeling Multiscale Time-Series Modeling
Advanced Computing, Mathematics and Data Division, Pacific Northwest National Laboratory, WA 99354 · Massachusetts Institute of Technology, MA 02139 · Oden Institute for Computational Engineering and Sciences, University of Texas at Austin
Privacy for Free in Efficient Agnostic Halfspace Learning
Nov 3, 2026Giorgio Racca, Amartya Sanyal, Vikrant SinghalHalfspace Learning Agnostic Learning
Department of Computer Science, University of Copenhagen

Nov 3, 2026J/K move · Enter open · S save

DreamX: Video Diffusion Model is a Universal Latent Action Learner

Jisoo Kim, Chaehyun Kim, Junsu Kim, Jiho Park+5

KAIST · NVIDIA

Latent Action Learning Action Representation Learning Cross-Embodiment Transfer Video Diffusion Models

Learning a universal action representation from video requires modeling dynamics that are consistent across different executions of the same action yet invariant to changes in embodiment and appearance. Existing latent action models infer actions from visual transitions, but their representations often entangle the underlying action with irrelevant visual variations, limiting transfer across embodiments. We introduce DreamX, which repurposes a pretrained video diffusion model as a universal latent action learner. Our approach extracts compact latent actions from the video model's rich spatiotemporal features, compressing each latent video frame by

\boldsymbol{300\times}

(

\boldsymbol{1{,}200 \rightarrow 4}

tokens). In addition, by leveraging the video model's generative prior, these latent actions capture action-relevant dynamics disentangled from visual appearance. The resulting latent space serves as a shared action interface across embodiments (human and robot), environments, and action parameterizations. On LARY-Bench, DreamX achieves state-of-the-art performance among latent action models. Moreover, actions encoded in one setting remain effective in another: the latent space supports any-to-any video transfer and zero-shot action transfer across diverse embodiments and environments. Furthermore, this shared interface also enables policy learning in a compact latent action space, reducing the computational cost of video-diffusion-based policies.