Endless Terminals: Scaling RL Environments for Terminal Agents
Paper • 2601.16443 • Published • 18
Reading
-
-
Linear representations in language models can change dramatically over a conversation
Paper • 2601.20834 • Published • 21 -
Scaling Embeddings Outperforms Scaling Experts in Language Models
Paper • 2601.21204 • Published • 102 -
Teaching Models to Teach Themselves: Reasoning at the Edge of Learnability
Paper • 2601.18778 • Published • 41 -
DeepSearchQA: Bridging the Comprehensiveness Gap for Deep Research Agents
Paper • 2601.20975 • Published • 10 -
Retrieval-Infused Reasoning Sandbox: A Benchmark for Decoupling Retrieval and Reasoning Capabilities
Paper • 2601.21937 • Published • 19 -
Length-Unbiased Sequence Policy Optimization: Revealing and Controlling Response Length Variation in RLVR
Paper • 2602.05261 • Published • 52 -
EgoActor: Grounding Task Planning into Spatial-aware Egocentric Actions for Humanoid Robots via Visual-Language Models
Paper • 2602.04515 • Published • 39 -
F-GRPO: Don't Let Your Policy Learn the Obvious and Forget the Rare
Paper • 2602.06717 • Published • 74 -
OPUS: Towards Efficient and Principled Data Selection in Large Language Model Pre-training in Every Iteration
Paper • 2602.05400 • Published • 349 -
PhyCritic: Multimodal Critic Models for Physical AI
Paper • 2602.11124 • Published • 54 -
Composition-RL: Compose Your Verifiable Prompts for Reinforcement Learning of Large Language Models
Paper • 2602.12036 • Published • 93 -
Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments
Paper • 2602.11964 • Published • 12 -
GEBench: Benchmarking Image Generation Models as GUI Environments
Paper • 2602.09007 • Published • 39 -
GLM-5: from Vibe Coding to Agentic Engineering
Paper • 2602.15763 • Published • 120 -
DeepImageSearch: Benchmarking Multimodal Agents for Context-Aware Image Retrieval in Visual Histories
Paper • 2602.10809 • Published • 58 -
Experiential Reinforcement Learning
Paper • 2602.13949 • Published • 71 -
REDSearcher: A Scalable and Cost-Efficient Framework for Long-Horizon Search Agents
Paper • 2602.14234 • Published • 26 -
ResearchGym: Evaluating Language Model Agents on Real-World AI Research
Paper • 2602.15112 • Published • 21 -
SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks
Paper • 2602.12670 • Published • 56 -
jina-embeddings-v5-text: Task-Targeted Embedding Distillation
Paper • 2602.15547 • Published • 26 -
Discovering Multiagent Learning Algorithms with Large Language Models
Paper • 2602.16928 • Published • 16 -
Towards a Science of AI Agent Reliability
Paper • 2602.16666 • Published • 15 -
Empty Shelves or Lost Keys? Recall Is the Bottleneck for Parametric Factuality
Paper • 2602.14080 • Published • 20 -
A Very Big Video Reasoning Suite
Paper • 2602.20159 • Published • 517 -
TOPReward: Token Probabilities as Hidden Zero-Shot Rewards for Robotics
Paper • 2602.19313 • Published • 25 -
Mobile-O: Unified Multimodal Understanding and Generation on Mobile Device
Paper • 2602.20161 • Published • 23 -
Decoding as Optimisation on the Probability Simplex: From Top-K to Top-P (Nucleus) to Best-of-K Samplers
Paper • 2602.18292 • Published • 11 -
Unified Latents (UL): How to train your latents
Paper • 2602.17270 • Published • 58 -
DODO: Discrete OCR Diffusion Models
Paper • 2602.16872 • Published • 9 -
Agentic Code Reasoning
Paper • 2603.01896 • Published • 10 -
CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation
Paper • 2602.24286 • Published • 97 -
Learning When to Act or Refuse: Guarding Agentic Reasoning Models for Safe Multi-Step Tool Use
Paper • 2603.03205 • Published • 12 -
Beyond Language Modeling: An Exploration of Multimodal Pretraining
Paper • 2603.03276 • Published • 102 -
Memex(RL): Scaling Long-Horizon LLM Agents via Indexed Experience Memory
Paper • 2603.04257 • Published • 19 -
DARE: Aligning LLM Agents with the R Statistical Ecosystem via Distribution-Aware Retrieval
Paper • 2603.04743 • Published • 52 -
Believe Your Model: Distribution-Guided Confidence Calibration
Paper • 2603.03872 • Published • 40 -
Lost in Stories: Consistency Bugs in Long Story Generation by LLMs
Paper • 2603.05890 • Published • 91 -
Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMs
Paper • 2603.09906 • Published • 75 -
Recursive Language Models Meet Uncertainty: The Surprising Effectiveness of Self-Reflective Program Search for Long Context
Paper • 2603.15653 • Published • 12 -
ProRL Agent: Rollout-as-a-Service for RL Training of Multi-Turn LLM Agents
Paper • 2603.18815 • Published • 14 -
Hyperagents
Paper • 2603.19461 • Published • 43 -
On the Direction of RLVR Updates for LLM Reasoning: Identification and Exploitation
Paper • 2603.22117 • Published • 28 -
Reasoning as Compression: Unifying Budget Forcing via the Conditional Information Bottleneck
Paper • 2603.08462 • Published • 21 -
MinerU-Diffusion: Rethinking Document OCR as Inverse Rendering via Diffusion Decoding
Paper • 2603.22458 • Published • 131 -
From Static Templates to Dynamic Runtime Graphs: A Survey of Workflow Optimization for LLM Agents
Paper • 2603.22386 • Published • 54 -
Ego2Web: A Web Agent Benchmark Grounded in Egocentric Videos
Paper • 2603.22529 • Published • 6 -
UI-Voyager: A Self-Evolving GUI Agent Learning via Failed Experience
Paper • 2603.24533 • Published • 44 -
Composer 2 Technical Report
Paper • 2603.24477 • Published • 14