| Online Skill Learning for Web Agents via State-Grounded Dynamic Retrieval | Agents | Jiaxi Li +7 | Jun 3, 2026 | 10 |
| SkillHarm: Lifecycle-Aware Skill-Based Attacks via Automated Construction | Agents | Yuting Ning +10 | Jun 1, 2026 | 10 |
| Light-WAM: Efficient World Action Models with State-Fusion Action Decoding | Robotics | Ziang Li +7 | Jun 6, 2026 | 10 |
| Why Muon Outperforms Adam: A Curvature Perspective | Safety & alignment | Shuche Wang +4 | Jun 3, 2026 | 10 |
| Text-to-Image Models Need Less from Text Encoders Than You Think | Image & video gen | Nurit Spingarn +3 | Jun 2, 2026 | 10 |
| AffordanceVLA: A Vision-Language-Action Model Empowering Action Generation through Affordance-Aware Understanding | Robotics | Qize Yu +12 | Jun 4, 2026 | 10 |
| MAOAM: Unified Object and Material Selection with Vision-Language Models | Image & video gen | Jaden Park +7 | Jun 2, 2026 | 10 |
| Large Language Models Hack Rewards, and Society | Reinforcement learning | Wei Liu +4 | Jun 2, 2026 | 10 |
| Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models | Vision & multimodal | Glenn Jocher +5 | Jun 2, 2026 | 10 |
| Agent Skills Should Go Beyond Text: The Case for Visual Skills | Vision & multimodal | Binxiao Xu +3 | May 31, 2026 | 10 |
| RoboStressBench: Benchmarking VLM Robustness to Physical Visual Stress in Embodied Scenes | Robotics | Leyi Wu +13 | May 30, 2026 | 10 |
| Speculative Pipeline Decoding: Higher-Accruacy and Zero-Bubble Speculation via Pipeline Parallelism | Code | Yijiong Yu +4 | May 29, 2026 | 10 |
| PARCEL: Pool-Anchored Resampling with Conditioned Elastic Queries for Efficient Vision-Language Understanding | Vision & multimodal | Selim Kuzucu +5 | May 28, 2026 | 10 |
| The Flip Side of RLHF: On-Policy Feedback for Reward Model Self-Supervised Improvement | Reinforcement learning | Xiaobo Wang +5 | May 29, 2026 | 10 |
| VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies | Robotics | Mingjian Gao +11 | May 28, 2026 | 10 |
| REPOT: Recoverable Program-of-Thought via Checkpoint Repair | Data & benchmarks | Parsa Mazaheri | May 28, 2026 | 10 |
| Verifiable Rewards Beyond Math and Code: Lightweight Corpus-Grounded Process Supervision for Factual Question Answering | Code | Shicheng Fan +5 | May 28, 2026 | 10 |
| ChildVox: A Speech, Audio, and Large Audio-Language Model Benchmark in Understanding and Characterizing Sound across Childhood | Speech & audio | Tiantian Feng +12 | May 28, 2026 | 10 |
| Thinking Before Constraining: A Unified Decoding Framework for Large Language Models | Code | Ngoc Trinh Hung Nguyen +5 | May 28, 2026 | 10 |
| Category-Level 3D Correspondence in Camera Space via Morphable Object Priors | Robotics | Leonhard Sommer +3 | May 27, 2026 | 10 |
| Rethinking How to Remember: Beyond Atomic Facts in Lifelong LLM Agent Memory | Agents | Jingwei Sun +4 | May 19, 2026 | 10 |
| WildCity: A Real-World City-Scale Testbed for Rendering, Simulation, and Spatial Intelligence | Robotics | Xiangyu Han +11 | Jul 7, 2026 | 9 |
| CGGS: Consistency-Augmented Geometric Gaussian Splatting for Ego-centric 3D Scene Generation | Image & video gen | Zhenyu Sun +3 | Jul 4, 2026 | 9 |
| Deform360: A Massive Multi-view Visuotactile Dataset for Deformable World Models | Robotics | Hongyu Li +13 | Jul 6, 2026 | 9 |
| PraMem: Practice-derived Experiential Memory for Long-horizon Behavior Prediction | Code | Zhuoqun Li +13 | Jul 3, 2026 | 9 |
| Teaching LLMs to Recommend and Defer in Underrepresented Epilepsy Care | Agents | Shreyas Rajesh +10 | Jun 30, 2026 | 9 |
| WARP: Weight-Space Analysis for Recovering Training Data Portfolios | Efficiency & systems | Tzu-Heng Huang +3 | Jul 2, 2026 | 9 |
| AnyGroundBench: A Specialized-Domain Benchmark for Video Grounding in Vision-Language Models | Vision & multimodal | Rintaro Otsubo +8 | Jul 2, 2026 | 9 |
| NoPA: Non-Parametric Online 3D Scene Graph Generation | Efficiency & systems | Qi Xun Yeo +3 | Jul 1, 2026 | 9 |
| AtomiMed: Hierarchical Atomic Fact-Checking for Universal Clinical-Aware Medical Report Evaluation | Speech & audio | Yuan Wang +10 | Jun 30, 2026 | 9 |
| Goku: A Million-Scale Universal Dataset and Benchmark for Instruction-Based Video Editing | Robotics | Sen Liang +10 | Jun 30, 2026 | 9 |
| AVTok: 1D Unified Tokenization for Holistic Audio-Video Generation | Speech & audio | Kien T. Pham +3 | Jun 29, 2026 | 9 |
| Cognitive Episodes in LLM Reasoning Traces Enable Interpretable Human Item Difficulty Prediction | Safety & alignment | Chenguang Wang +6 | Jun 26, 2026 | 9 |
| Hallucination in World Models is Predictable and Preventable | Image & video gen | Nicklas Hansen +1 | Jun 25, 2026 | 9 |
| CoffeeBench: Benchmarking Long-Horizon LLM Agents in Heterogeneous Multi-Agent Economies | Agents | Issa Sugiura +7 | Jun 15, 2026 | 9 |
| Look Light, Think Heavy: What Multimodal Chain-of-Thought Reasoning Can and Cannot Do | Vision & multimodal | Zhuoran Jin +7 | Jun 21, 2026 | 9 |
| Tapered Language Models | Data & benchmarks | Reza Bayat +2 | Jun 22, 2026 | 9 |
| Rethinking Shrinkage Bias in LLM FP4 Pretraining: Geometric Origin, Systemic Impact, and UFP4 Recipe | Safety & alignment | Qian Zhao +11 | Jun 18, 2026 | 9 |
| Understanding the Behaviors of Environment-aware Information Retrieval | Code | Ruifeng Yuan +6 | Jun 15, 2026 | 9 |
| ViT-Up: Faithful Feature Upsampling for Vision Transformers | Vision & multimodal | Krispin Wandel +2 | Jun 12, 2026 | 9 |
| ProCUA-SFT Technical Report | Vision & multimodal | Jaehun Jung +13 | Jun 15, 2026 | 9 |
| RefGC-SR^2: Reference-guided Generated Content Super-Resolution and Refinement | Image & video gen | Jeahun Sung +3 | Jun 13, 2026 | 9 |
| Prompt-Level Distillation: A Non-Parametric Alternative to Model Fine-Tuning for Efficient Reasoning | Safety & alignment | Sanket Badhe +1 | Jun 2, 2026 | 9 |
| The Hidden Power of Scaling Factor in LoRA Optimization | Efficiency & systems | Zicheng Zhang +12 | Jun 11, 2026 | 9 |
| Avatar V: Scaling Video-Reference Avatar Video Generation | Image & video gen | Benjamin Liang +22 | Jun 11, 2026 | 9 |
| Risk Under Pressure: Compute-Aware Evaluation of Adversarial Robustness in Language Models | Speech & audio | Malikeh Ehghaghi +3 | Jun 9, 2026 | 9 |
| OPRD: On-Policy Representation Distillation | Code | Shenzhi Yang +9 | Jun 4, 2026 | 9 |
| PlatonicNav: Unveiling Semantic Correspondence in Navigation with Platonic Topological Maps | Robotics | Junlin Long +7 | Jun 1, 2026 | 9 |
| Diagnosing Harmful Continuation in Answer-Correct Long-CoT Training Traces | Safety & alignment | Chen He +4 | May 28, 2026 | 9 |
| Hide-and-Seek in Trajectories: Discovering Failure Signals for VLA Runtime Monitoring | Robotics | Seongheon Park +6 | May 29, 2026 | 9 |
| From Model Scaling to System Scaling: Scaling the Harness in Agentic AI | Agents | Shangding Gu | May 25, 2026 | 9 |
| CoHyDE: Iterative Co-Training of LLM Rewriter & Dense Encoder for Tool Retrieval | Agents | Vaishali Senthil +2 | May 28, 2026 | 9 |
| CONF-KV: Confidence-Aware KV Cache Eviction with Mixed-Precision Storage for Long-Horizon LLM | Vision & multimodal | Yubo Li +1 | May 24, 2026 | 9 |
| Reflective Prompt Tuning through Language Model Function-Calling | Efficiency & systems | Farima Fatahi Bayat +3 | May 20, 2026 | 9 |
| SAM: State-Adaptive Memory for Long-Horizon Reasoning Agent | Agents | Yuyang Hu +7 | May 23, 2026 | 9 |
| Beyond Final Answers: Auditing Trajectory-Level Hallucinations in Multi-Agent Industrial Workflows | Vision & multimodal | Harshada Badave +8 | May 22, 2026 | 9 |
| CoSPlay: Cooperative Self-Play at Test-Time with Self-Generated Code and Unit Test | Speech & audio | Zhangyi Hu +8 | May 22, 2026 | 9 |
| OPSD-V: On-Policy Self-Distillation for Post-Training Few-Step Autoregressive Video Generators | Image & video gen | Hongyu Liu +8 | Jul 9, 2026 | 8 |
| TREK: Distill to Explore, Reinforce to Refine | Agents | Yuanda Xu +12 | Jul 6, 2026 | 8 |
| Attending to Multimodal Generation One Token at a Time | Vision & multimodal | Varun Gupta +2 | Jul 4, 2026 | 8 |
| Mastermind: Strategy-grounded Learning for Repository-Scale Vulnerability Reproduction | Agents | Mingzhe Du +6 | Jul 2, 2026 | 8 |
| Are Performance-Optimization Benchmarks Reliably Measuring Coding Agents? | Agents | Zhi Chen +4 | Jul 1, 2026 | 8 |
| Autonomous Scientific Discovery via Iterative Meta-Reflection | Vision & multimodal | Bingchen Zhao +2 | Jul 1, 2026 | 8 |
| GRPO, Dr. GRPO, and DAPO Are Three Operations on One Number: The Group-Standard-Deviation Identity | Reinforcement learning | Yong Yi Bay +1 | Jun 30, 2026 | 8 |
| SciIR: A Large-scale Training Dataset and Benchmark for Scientific Image Reasoning Generation | Image & video gen | Zhiyuan Ma +8 | Jun 29, 2026 | 8 |
| When More Sampling Hurts: The Modal Ceiling and Correlation Ceiling of Test-Time Scaling | Data & benchmarks | Yong Yi Bay +1 | Jun 27, 2026 | 8 |
| Building to the Test: Coding Agents Deliver What You Check, Not What You Requested | Agents | Yanuo Ma +2 | Jun 26, 2026 | 8 |
| AI translation of literary texts is "fine", but readers still prefer human translations | Agents | Yves Ferstler +5 | Jun 24, 2026 | 8 |
| One Forward Beats Two: InnerZoom for Accurate and Efficient GUI Grounding | Vision & multimodal | Chen Liu +7 | Jun 29, 2026 | 8 |
| Walking in the Implicit: Interactive World Exploration via Neural Scene Representation | Image & video gen | Zhiqi Li +8 | Jun 29, 2026 | 8 |
| MirrorPPR: Exemplar-Based Portrait Photo Retouching | Image & video gen | Zhihong Liu +6 | Jun 28, 2026 | 8 |
| PolicyGuard: A Dialogue-Grounded Sub-Agent Verifier for Policy Adherence in LLM Agents | Agents | Seongjae Kang +2 | Jun 28, 2026 | 8 |
| Ko-WideSearch: A Korean Breadth-Search Benchmark for Exhaustive Set Enumeration by Web Agents | Agents | Minbyul Jeong | Jun 25, 2026 | 8 |
| AgentOdyssey: Open-Ended Long-Horizon Text Game Generation for Test-Time Continual Learning Agents | Agents | Zheyuan Zhang +6 | May 29, 2026 | 8 |
| RoPE-Aware Bit Allocation for KV-Cache Quantization | Code | Fengfeng Liang +2 | Jun 23, 2026 | 8 |
| Safe Few-Step Generation via Velocity Editing | Robotics | Yujin Choi +1 | Jun 22, 2026 | 8 |
| Connect the Dots: Training LLMs for Long-Lifecycle Agents with Cross-Domain Generalization Via Reinforcement Learning | Agents | Yanxi Chen +6 | Jun 18, 2026 | 8 |
| Freeing the Law with LOCUS: A Local Ordinance Corpus for the United States | Vision & multimodal | Denis Peskoff +3 | Jun 17, 2026 | 8 |
| LooseControlVideo: Directorial Video Control using Spatial Blocking | Image & video gen | Shariq Farooq Bhat +2 | Jun 17, 2026 | 8 |
| CEO-Bench: Can Agents Play the Long Game? | Speech & audio | Haozhe Chen +2 | Jun 16, 2026 | 8 |
| SciOrch: Learning to Orchestrate Expert LLMs for Solving Frontier Multimodal Scientific Reasoning Tasks | Vision & multimodal | Jingru Guo +8 | Jun 14, 2026 | 8 |
| ActWorld: From Explorable to Interactive World Model via Action-Aware Memory | Image & video gen | Zhexiao Xiong +13 | Jun 16, 2026 | 8 |
| ChLogic: Evaluating Robustness of Logical Reasoning in Chinese Expressions | Data & benchmarks | Peixian Zhou +5 | Jun 16, 2026 | 8 |
| A Gradient Perspective on RLVR Stability and Winner Advantage Policy Optimization | Code | Prasanth YSS +8 | Jun 15, 2026 | 8 |
| Unstable Features, Reproducible Subspaces: Understanding Seed Dependence in Sparse Autoencoders | Efficiency & systems | Gleb Gerasimov +5 | Jun 10, 2026 | 8 |
| VISTA: View-Consistent Self-Verified Training for GUI Grounding | Reinforcement learning | Xinyu Qiu +5 | Jun 12, 2026 | 8 |
| ClinHallu: A Benchmark for Diagnosing Stage-Wise Hallucinations in Medical MLLM Reasoning | Vision & multimodal | Sicheng Yang +7 | Jun 12, 2026 | 8 |
| When is Your LLM Steerable? | Code | Chenrui Fan +4 | Jun 10, 2026 | 8 |
| Breaking the Bubble: Asynchronous Pipeline Parallel Training with Bounded Weight Inconsistency | Efficiency & systems | Itay Elam +3 | Jun 5, 2026 | 8 |
| BrainSurgery: Reproducible and Reliable Declarative Weight Manipulations for Model Editing and Upcycling | Robotics | Gianluca Barmina +4 | Jun 8, 2026 | 8 |
| Reasoning Arena: Trace Tournaments When Verifiable Rewards Fall Short | Code | Han Zhou +4 | Jun 8, 2026 | 8 |
| Answer Presence Drives RAG Rewriting Gains | Code | Yuejie Li +10 | Jun 4, 2026 | 8 |
| Liberating LLM Capabilities in Full-Duplex Speech Models | Speech & audio | Luoyuan Zhang +6 | May 4, 2026 | 8 |
| LLMs Can Leak Training Data But Do They Want To? A Propensity-Aware Evaluation of Memorization in LLMs | Data & benchmarks | Gianluca Barmina +2 | Jun 4, 2026 | 8 |
| World-Language-Action Model for Unified World Modeling, Language Reasoning, and Action Synthesis | Robotics | Yi Yang +11 | Jun 4, 2026 | 8 |
| WebRISE: Requirement-Induced State Evaluation for MLLM-Generated Web Artifacts | Vision & multimodal | Yuxin Meng +11 | Jun 2, 2026 | 8 |
| Economy of Minds: Emerging Multi-Agent Intelligence with Economic Interactions | Agents | Zhenting Qi +15 | Jun 1, 2026 | 8 |
| Value-Aware Stochastic KV Cache Eviction for Reasoning Models | Efficiency & systems | Ting-Yun Chang +5 | Jun 2, 2026 | 8 |
| OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification | Reinforcement learning | Yuhang Zhou +7 | May 31, 2026 | 8 |
| MERIT: Learning Disentangled Music Representations for Audio Similarity | Speech & audio | Abhinaba Roy +2 | May 26, 2026 | 8 |