mlsanityData to 14 Oct 2026

WebSailor-V2: Bridging the Chasm to Proprietary Agents via Synthetic Data and Scalable Reinforcement Learning

arXiv 2509.13305 ↗First posted 16 Sep 2025cs.LG cs.CLcited by 61

On the map: LLM agents and tool use → Web, search, and research agents

also in 2 more
Abstract (arXiv)

Transcending human cognitive limitations represents a critical frontier in LLM training. Proprietary agentic systems like DeepResearch have demonstrated superhuman capabilities on extremely complex information-seeking benchmarks such as BrowseComp, a feat previously unattainable. We posit that their success hinges on a sophisticated reasoning pattern absent in open-source models: the ability to systematically reduce extreme uncertainty when navigating vast information landscapes. Based on this insight, we introduce WebSailor, a complete post-training methodology designed to instill this crucial capability. Our approach involves generating novel, high-uncertainty tasks through structured sampling and information obfuscation, RFT cold start, and an efficient agentic RL training algorithm, Duplicating Sampling Policy Optimization (DUPO). With this integrated pipeline, WebSailor significantly outperforms all open-source agents in complex information-seeking tasks, matching proprietary agents' performance and closing the capability gap.

Key facts. Cited by 61 papers in the map (data through 14 Oct 2026). Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use. Builds on 39 of its 55 references that are in the map. First posted on arXiv 16 Sep 2025.

Builds on

39 of 55 references.

Closest ancestors · 6

Same direction, newest first

  1. 2025DeepDive: Advancing Deep Search Agents with Knowledge Graphs and Multi-Turn RL12 Sep
  2. 2025WebExplorer: Explore and Evolve for Training Long-Horizon Web Agents8 Sep
  3. 2025Beyond Ten Turns: Unlocking Long-Horizon Agentic Search with Large-Scale Asynchronous RL11 Aug
  4. 2025WebSailor: Navigating Super-human Reasoning for Web Agent3 Jul
  5. 2025DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents13 Jun
  6. 2025WebDancer: Towards Autonomous Information Seeking Agency28 May

Other references · 33

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  2. 2025Qwen3 Technical Reportcited by 8,820
  3. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  4. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  5. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scalecited by 2,810
  6. 2019Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelismcited by 1,868
  7. 2025Understanding R1-Zero-Like Training: A Critical Perspectivecited by 1,345
  8. 2023GAIA: a benchmark for General AI Assistantscited by 1,128
  9. 2025Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?cited by 982
  10. 2025BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agentscited by 647
  11. 2025UI-TARS: Pioneering Automated GUI Interaction with Native Agentscited by 610
  12. 2025Humanity's Last Examcited by 491
21 more
  1. 2025GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Modelscited by 467
  2. 2025WebThinker: Empowering Large Reasoning Models with Deep Research Capabilitycited by 335
  3. 2025MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agentcited by 246
  4. 2025MEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon Agentscited by 238
  5. 2025UI-TARS-2 Technical Report: Advancing GUI Agent with Multi-Turn Reinforcement Learningcited by 215
  6. 2025WebWalker: Benchmarking LLMs in Web Traversalcited by 175
  7. 2025BrowseComp-ZH: Benchmarking Web Browsing Ability of Large Language Models in Chinesecited by 145
  8. 2025OWL: Optimized Workforce Learning for General Multi-Agent Assistance in Real-World Task Automationcited by 131
  9. 2024Leave No Document Behind: Benchmarking Long-Context LLMs with Extended Multi-Doc QAcited by 122
  10. 2025WebShaper: Agentically Data Synthesizing via Information-Seeking Formalizationcited by 121
  11. 2025WebWatcher: Breaking New Frontier of Vision-Language Deep Research Agentcited by 119
  12. 2025Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolutioncited by 116
  13. 2025Reinforcement Learning for Long-Horizon Interactive LLM Agentscited by 108
  14. 2020RetinaGAN: An Object-aware Approach to Sim-to-Real Transfercited by 70
  15. 2023Calibrating LLM-Based Evaluatorcited by 53
  16. 2019Simulation-based reinforcement learning for real-world autonomous drivingcited by 52
  17. 2025SciMaster: Towards General-Purpose Scientific AI Agents, Part I. X-Master as Foundation: Can We Lead on Humanity's Last Exam?cited by 47
  18. 2025A Survey of Sim-to-Real Methods in RL: Progress, Prospects and Challenges with Foundation Modelscited by 38
  19. 2025SFR-DeepResearch: Towards Effective Reinforcement Learning for Autonomously Reasoning Single Agentscited by 31
  20. 2025LaRA: Benchmarking Retrieval-Augmented Generation and Long-Context LLMs -- No Silver Bullet for LC or RAG Routingcited by 29
  21. 2025DeepDiver: Adaptive Search Intensity Scaling via Open-Web Reinforcement Learningcited by 23

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 16; 2026: 45.

61 papers.

Sort:

By area:

  1. 2025Scaling Long-Horizon LLM Agent via Context-Foldingcited by 148
  2. 2025AgentFold: Long-Horizon Web Agents with Proactive Context Managementcited by 103
  3. 2025DeepAgent: A General Reasoning Agent with Scalable Toolsetscited by 71
  4. 2025Tongyi DeepResearch Technical Reportcited by 66
  5. 2026REDSearcher: A Scalable and Cost-Efficient Framework for Long-Horizon Search Agentscited by 49
  6. 2026OpenSeeker: Democratizing Frontier Search Agents by Fully Open-Sourcing Training Datacited by 39
  7. 2025Search Self-play: Pushing the Frontier of Agent Capability without Supervisioncited by 34
  8. 2025A Survey on Agentic Multimodal Large Language Modelscited by 23
  9. 2025A Comprehensive Survey on Reinforcement Learning-based Agentic Search: Foundations, Roles, Optimizations, Evaluations, and Applicationscited by 22
  10. 2025WebLeaper: Empowering Efficiency and Efficacy in WebAgent via Enabling Info-Rich Seekingcited by 18
  11. 2025Toward Effective Tool-Integrated Reasoning via Self-Evolved Preference Learningcited by 17
  12. 2026Reasoning and Tool-use Compete in Agentic RL:From Quantifying Interference to Disentangled Tuningcited by 17
38 more
  1. 2026The Confidence Dichotomy: Analyzing and Mitigating Miscalibration in Tool-Use Agentscited by 16
  2. 2025ParallelMuse: Agentic Parallel Thinking for Deep Information Seekingcited by 14
  3. 2026Chaining the Evidence: Robust Reinforcement Learning for Deep Search Agents with Citation-Aware Rubric Rewardscited by 14
  4. 2026DIVE: Scaling Diversity in Agentic Task Synthesis for Generalizable Tool Usecited by 13
  5. 2026OffSeeker: Online Reinforcement Learning Is Not All You Need for Deep Research Agentscited by 12
  6. 2025Repurposing Synthetic Data for Fine-grained Search Agent Supervisioncited by 11
  7. 2025AgentProg: Empowering Long-Horizon GUI Agents with Program-Guided Context Managementcited by 11
  8. 2026O-Researcher: An Open Ended Deep Research Model via Multi-Agent Distillation and Agentic RLcited by 11
  9. 2026RE-TRAC: REcursive TRAjectory Compression for Deep Search Agentscited by 11
  10. 2026LiteResearcher: A Scalable Agentic RL Training Framework for Deep Research Agentcited by 10
  11. 2025Taxonomy, Evaluation and Exploitation of IPI-Centric LLM Agent Defense Frameworkscited by 9
  12. 2026AgentSwing: Adaptive Parallel Context Management Routing for Long-Horizon Web Agentscited by 9
  13. 2025EcomBench: Towards Holistic Evaluation of Foundation Agents in E-commercecited by 8
  14. 2026TRACE: Trajectory-Aware Comprehensive Evaluation for Deep Research Agentscited by 8
  15. 2026Demystifying Reinforcement Learning for Long-Horizon Tool-Using Agents: A Comprehensive Recipecited by 8
  16. 2026OpenSeeker-v2: Pushing the Limits of Search Agents with Informative and High-Difficulty Trajectoriescited by 8
  17. 2026Visual-Seeker: Towards Visual-Native Multimodal Agentic Search via Active Visual Reasoningcited by 8
  18. 2025InfoAgent: Advancing Autonomous Information-Seeking Agentscited by 7
  19. 2026CorpusQA: A 10 Million Token Benchmark for Corpus-Level Analysis and Reasoningcited by 7
  20. 2025Nested Browser-Use Learning for Agentic Information Seekingcited by 6
  21. 2026Argus: Evidence Assembly for Scalable Deep Research Agentscited by 6
  22. 2026DR-Venus: Towards Frontier Edge-Scale Deep Research Agents with Only 10K Open Datacited by 5
  23. 2026SearchEyes: Towards Frontier Multimodal Deep Search Intelligence via Search World Simulationcited by 5
  24. 2026OneDayAgent: Towards a Long-Horizon Harness for Autonomous Agentscited by 4
  25. 2026Beyond Accuracy: Unveiling Inefficiency Patterns in Tool-Integrated Reasoningcited by 3
  26. 2026LoHoSearch: Benchmarking Long-Horizon Search Agents Beyond the Human Difficulty Ceilingcited by 3
  27. 2026Agentic-Ideation: Sample Efficient Agentic Trajectories Synthesis for Scientific Ideation Agentscited by 3
  28. 2026Think Big, Search Small: Where Capacity Matters in Hierarchical Search Agents?cited by 3
  29. 2026SearchMaster: Grounded and Regulated Self-Play for Search Agentscited by 3
  30. 2026What Makes Good Agentic Data? An ACE Lens on Data Generation for LLM Agentscited by 3
  31. 2026Scalable Behaviour Cloning on Browser Using via Skill Distillationcited by 2
  32. 2026HyMem: Hierarchical Context Management for Long-Horizon Agents via Information Isolationcited by 2
  33. 2026ToolSelf: Unifying Task Execution and Self-Reconfiguration via Tool-Driven Emergent Adaptationcited by 1
  34. 2026QuarkMedSearch: A Long-Horizon Deep Search Agent for Exploring Medical Intelligencecited by 1
  35. 2026SciResearcher: Scaling Deep Research Agents for Frontier Scientific Reasoningcited by 1
  36. 2026Efficient Agentic Reasoning Through Self-Regulated Simulative Planningcited by 1
  37. 2026ToolAnchor: Anchoring Counterfactual Context to Boost Agentic Tool-use Capabilitycited by 1
  38. 2026SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaborationcited by 1

50 most cited of 61 All 61 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026WebFactory: Automated Compression of Foundational Language Intelligence into Grounded Web Agents
  2. 2025Rethinking the Design of Reinforcement Learning-Based Deep Research Agents
  3. 2024WebRL: Training LLM Web Agents via Self-Evolving Online Curriculum Reinforcement Learning
  4. 2026SAIL: Scientific Agentic Intelligence via a Science-Aware Loop
  5. 2025How to Train Your LLM Web Agent: A Statistical Diagnosis
  6. 2025WebResearcher: Unleashing unbounded reasoning capability in Long-Horizon Agents
  7. 2025DSPO: Stable and Efficient Policy Optimization for Agentic Search and Reasoning
  8. 2026From Proprietary to Open-Source: Bridging the Distribution Gap via Multi-Agent Protocol Distillation in Agentic Search