mlsanityData to 7 Oct 2026

MiroThinker-1.7 & H1: Towards Heavy-Duty Research Agents via Verification

arXiv 2603.15726 ↗First posted 16 Mar 2026cs.CL cs.AI cs.IRcited by 28

On the map: LLM reasoning → Knowledge- and tool-augmented reasoning

also in 1 more
Abstract (arXiv)

We present MiroThinker-1.7, a new research agent designed for complex long-horizon reasoning tasks. Building on this foundation, we further introduce MiroThinker-H1, which extends the agent with heavy-duty reasoning capabilities for more reliable multi-step problem solving. In particular, MiroThinker-1.7 improves the reliability of each interaction step through an agentic mid-training stage that emphasizes structured planning, contextual reasoning, and tool interaction. This enables more effective multi-step interaction and sustained reasoning across complex tasks. MiroThinker-H1 further incorporates verification directly into the reasoning process at both local and global levels. Intermediate reasoning decisions can be evaluated and refined during inference, while the overall reasoning trajectory is audited to ensure that final answers are supported by coherent chains of evidence. Across benchmarks covering open-web research, scientific reasoning, and financial analysis, MiroThinker-H1 achieves state-of-the-art performance on deep research tasks while maintaining strong results on specialized domains. We also release MiroThinker-1.7 and MiroThinker-1.7-mini as open-source models, providing competitive research-agent capabilities with significantly improved efficiency.

Key facts. Cited by 28 papers in the map (data through 7 Oct 2026). Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning. Builds on 26 of its 39 references that are in the map. First posted on arXiv 16 Mar 2026.

Builds on

26 of 39 references.

Closest ancestors · 4

Same direction, newest first

  1. 2026FrontierScience: Evaluating AI's Ability to Perform Expert-Level Scientific Tasks29 Jan
  2. 2025SealQA: Raising the Bar for Reasoning in Search-Augmented Language Models1 Jun
  3. 2025WebThinker: Empowering Large Reasoning Models with Deep Research Capability30 Apr
  4. 2024Fact, Fetch, and Reason: A Unified Evaluation of Retrieval-Augmented Generation19 Sep

Other references · 22

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  2. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  3. 2025Qwen3 Technical Reportcited by 8,195
  4. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  5. 2023GAIA: a benchmark for General AI Assistantscited by 1,084
  6. 2025BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agentscited by 616
  7. 2026GLM-5: from Vibe Coding to Agentic Engineeringcited by 502
  8. 2025DeepResearcher: Scaling Deep Research via Reinforcement Learning in Real-world Environmentscited by 268
  9. 2024Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimizationcited by 238
  10. 2025WebWalker: Benchmarking LLMs in Web Traversalcited by 171
  11. 2025BrowseComp-ZH: Benchmarking Web Browsing Ability of Large Language Models in Chinesecited by 139
  12. 2025xbench: Tracking Agents Productivity Scaling with Profession-Aligned Real-World Evaluationscited by 118
10 more
  1. 2024Provably Mitigating Overoptimization in RLHF: Your SFT Loss is Implicitly an Adversarial Regularizercited by 109
  2. 2025MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scalingcited by 72
  3. 2026DeepSearchQA: Bridging the Comprehensiveness Gap for Deep Research Agentscited by 44
  4. 2026REDSearcher: A Scalable and Cost-Efficient Framework for Long-Horizon Search Agentscited by 43
  5. 2025Scaling Agents via Continual Pre-trainingcited by 39
  6. 2025MedBrowseComp: Benchmarking Medical Deep Research and Computer Usecited by 28
  7. 2025FinSearchComp: Towards a Realistic, Expert-Level Evaluation of Financial Search and Reasoningcited by 26
  8. 2026DeepResearchEval: An Automated Framework for Deep Research Task Construction and Agentic Evaluationcited by 13
  9. 2026MiroFlow: Towards High-Performance and Robust Open-Source Agent Framework for General Deep Research Taskscited by 13
  10. 2025Step-DeepResearch Technical Reportcited by 12

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 28.

28 papers.

Sort:

By area:

  1. 2026Marco DeepResearch: Unlocking Efficient Deep Research Agents via Verification-Centric Designcited by 12
  2. 2026Towards Long-horizon Agentic Multimodal Searchcited by 9
  3. 2026LongSeeker: Elastic Context Orchestration for Long-Horizon Search Agentscited by 8
  4. 2026OpenSeeker-v2: Pushing the Limits of Search Agents with Informative and High-Difficulty Trajectoriescited by 7
  5. 2026PBSD: Privileged Bayesian Self-Distillation for Long-Horizon Credit Assignmentcited by 6
  6. 2026MiroEval: Benchmarking Multimodal Deep Research Agents in Process and Outcomecited by 5
  7. 2026SWE-AGILE: A Software Agent Framework for Efficiently Managing Dynamic Reasoning Contextcited by 5
  8. 2026Argus: Evidence Assembly for Scalable Deep Research Agentscited by 4
  9. 2026FORT-Searcher: Synthesizing Shortcut-Resistant Search Tasks for Training Deep Search Agentscited by 4
  10. 2026AREX: Towards a Recursively Self-Improving Agent for Deep Researchcited by 4
  11. 2026SearchSwarm: Towards Delegation Intelligence in Agentic LLMs for Long-Horizon Deep Researchcited by 3
  12. 2026S1-DeepResearch: Beyond Search, Toward Real-World Long-Horizon Research Agentscited by 3
16 more
  1. 2026Deep Research as Rubric for Reinforcement Learningcited by 2
  2. 2026DuMate-DeepResearch: An Auditable Multi-Agent System with Recursive Search and Rubric-Grounded Reasoningcited by 2
  3. 2026ACE: Pluggable Adaptive Context Elasticizer across Agentscited by 2
  4. 2026SearchArt: Training Long-Horizon Search Agent with Scalable Synthetic and Verified Taskcited by 2
  5. 2026Workflow Closure Is Not Scientific Closure in Auto-Research Systemscited by 1
  6. 2026Lean4Agent: Formal Modeling and Verification for Agent Workflow and Trajectorycited by 1
  7. 2026Are We Ready for AI-Driven Discovery? AI Verification Before the Next Fundamental Physics Breakthroughcited by 1
  8. 2026OneDayAgent: Towards a Long-Horizon Harness for Autonomous Agentscited by 1
  9. 2026ABSeeker: Training Long-Horizon Search Agents via Answer-Backtracked Credit Assignmentcited by 1
  10. 2026BioMedArena: An Open-source Toolkit for Building and Evaluating Biomedical Deep Research Agentscited by 0
  11. 2026Distilling Temporal Search and Reasoning: Evolving LLMs for Future Prediction via Harness-Assisted Efficient Data Synthesiscited by 0
  12. 2026CHIME: Credit-Aware Hierarchical Memory Evolution for Long-Horizon Agentic Planningcited by 0
  13. 2026Iris: Climbing to the Search Frontiercited by 0
  14. 2026Boosting Deepresearch and LongContext Ability with Self-Generated Deepresearch Rollouts Tracescited by 0
  15. 2026Raven: The Harness of Harnesses for Composable Agentic Intelligencecited by 0
  16. 2026PrimeSeeker: Capability-Oriented Supervision for Deep Search Agentscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026ReThinker: Scientific Reasoning by Rethinking with Guided Reflection and Confidence Control
  2. 2025R2-KG: General-Purpose Dual-Agent Framework for Reliable Reasoning on Knowledge Graphs
  3. 2026Learning What to Investigate Next: Meta-Reasoning for Long-Horizon Research Agents
  4. 2025WebResearcher: Unleashing unbounded reasoning capability in Long-Horizon Agents
  5. 2025Eigen-1: Adaptive Multi-Agent Refinement with Monitor-Based RAG for Scientific Reasoning
  6. 2026EigentSearch-Q+: Enhancing Deep Research Agents with Structured Reasoning Tools
  7. 2025VeriThinker: Learning to Verify Makes Reasoning Model Efficient
  8. 2025WebAggregator: Enhancing Compositional Reasoning Capabilities of Deep Research Agent Foundation Models