mlsanityData to 7 Oct 2026

Mach-Mind-4-Flash Technical Report

arXiv 2607.09375 ↗First posted 10 Jul 2026cs.LG cs.CLcited by 4

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

also in 1 more
Abstract (arXiv)

We present Mach-Mind-4-Flash, a 35B-parameter Mixture-of-Experts (MoE) agentic model with 3B activated parameters. Through post-training optimization alone without scaling pre-training compute, the model achieves performance on par with or surpassing that of 100B-parameter-class models. By introducing scalable agentic interaction environments for large-scale reinforcement learning, the model attains significant performance gains on real-world application tasks. Our pipeline comprises three stages: (1) a unified RL/OPD training infrastructure with dynamic multi-teacher scheduling and operator-level acceleration, delivering 17% end-to-end training speedup; (2) multiple domain-specific RL experts trained in parallel across Reasoning, General, and Agent tracks, then fused into a single generalist via Multi-Teacher On-Policy Distillation (MOPD) -- a routed reverse-KL objective that eliminates the see-saw degradation of mixed-reward RL; (3) Hybrid Median-length Policy Optimization (HMPO), a single-stage token-efficiency method that compresses reasoning chains by 19--46% with ≤0.7 percentage-point accuracy loss. Mach-Mind-4-Flash scores 92.70 on AIME'26, 82.82 on IFBench, 80.74 on Behavioral-SafetyBench, 75.80 on BFCL-v4, 72.31 on BrowseComp-zh, and 84.20 on ClawBench -- leading or matching models with 10--30× its activated size at a fraction of the inference cost.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 64 of its 78 references that are in the map. First posted on arXiv 10 Jul 2026.

Builds on

64 of 78 references.

Closest ancestors · 5

Same direction, newest first

  1. 2026Nemotron 3 Super: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning14 Apr
  2. 2026Claw-Eval: Towards Trustworthy Evaluation of Autonomous Agents7 Apr
  3. 2026Step 3.5 Flash: Open Frontier-Level Intelligence with 11B Active Parameters11 Feb
  4. 2025Kimi K2: Open Agentic Intelligence28 Jul
  5. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains17 Jun

Other references · 59

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  4. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  6. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  7. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,072
  8. 2021On the Opportunities and Risks of Foundation Modelscited by 3,961
  9. 2017Deep reinforcement learning from human preferencescited by 3,846
  10. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,193
  11. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 2,950
  12. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,748
38 more
  1. 2019Fine-Tuning Language Models from Human Preferencescited by 2,124
  2. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 1,906
  3. 2021WebGPT: Browser-assisted question-answering with human feedbackcited by 1,722
  4. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,628
  5. 2023Gorilla: Large Language Model Connected with Massive APIscited by 1,302
  6. 2025Understanding R1-Zero-Like Training: A Critical Perspectivecited by 1,276
  7. 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,235
  8. 2025Kimi k1.5: Scaling Reinforcement Learning with LLMscited by 1,193
  9. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,143
  10. 2023Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversationscited by 1,124
  11. 2021Ethical and social risks of harm from Language Modelscited by 957
  12. 2024OpenHands: An Open Platform for AI Software Developers as Generalist Agentscited by 956
  13. 2024SimPO: Simple Preference Optimization with a Reference-Free Rewardcited by 951
  14. 2026DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligencecited by 877
  15. 2025DeepSeek-V3.2: Pushing the Frontier of Open Large Language Modelscited by 755
  16. 2025BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agentscited by 616
  17. 2024DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Servingcited by 595
  18. 2023API-Bank: A Comprehensive Benchmark for Tool-Augmented LLMscited by 574
  19. 2025τ2-Bench: Evaluating Conversational Agents in a Dual-Control Environmentcited by 558
  20. 2024ORPO: Monolithic Preference Optimization without Reference Modelcited by 556
  21. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Modelscited by 456
  22. 2026Kimi K2.5: Visual Agentic Intelligencecited by 449
  23. 2023LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compressioncited by 443
  24. 2024Better & Faster Large Language Models via Multi-token Predictioncited by 344
  25. 2024LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compressioncited by 343
  26. 2024Agent-SafetyBench: Evaluating the Safety of LLM Agentscited by 280
  27. 2024ToolSandbox: A Stateful, Conversational, Interactive Evaluation Benchmark for LLM Tool Use Capabilitiescited by 272
  28. 2026Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipecited by 262
  29. 2025MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attentioncited by 248
  30. 2025Generalizing Verifiable Instruction Followingcited by 178
  31. 2026MiMo-V2-Flash Technical Reportcited by 163
  32. 2025Skywork Open Reasoner 1 Technical Reportcited by 161
  33. 2025BrowseComp-ZH: Benchmarking Web Browsing Ability of Large Language Models in Chinesecited by 139
  34. 2023Beyond One-Preference-Fits-All Alignment: Multi-Objective Direct Preference Optimizationcited by 121
  35. 2025xbench: Tracking Agents Productivity Scaling with Profession-Aligned Real-World Evaluationscited by 118
  36. 2024Rule Based Rewards for Language Model Safetycited by 106
  37. 2025WritingBench: A Comprehensive Benchmark for Generative Writingcited by 95
  38. 2025The Art of Scaling Reinforcement Learning Compute for LLMscited by 91

50 most cited of 59 All 59 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026Every Coin Has Two Sides: On the Dual Nature of Generalization in On-Policy Distillation of Large Language Modelscited by 3
  2. 2026ME-VLM: A Unified VLM for Embodied Cognition and Agent Coordinationcited by 0
  3. 2026Rufus-Air: An Open LLM Post-Training Recipecited by 0
  4. 2026MMVistaReason: Toward Open-Data and Post-Training Recipes for Multimodal Reasoningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025INTELLECT-3: Technical Report
  2. 2025Introducing LongCat-Flash-Thinking: A Technical Report
  3. 2026LongCat-Flash-Thinking-2601 Technical Report
  4. 2026Structuring MoE Expert Selection for Agentic Reinforcement Learning
  5. 2026LongCat-Flash-Prover: Advancing Native Formal Reasoning via Agentic Tool-Integrated Reinforcement Learning
  6. 2026Phase-Aware Mixture of Experts for Agentic Reinforcement Learning
  7. 2025Towards Stable and Effective Reinforcement Learning for Mixture-of-Experts
  8. 2025Each Prompt Matters: Scaling Reinforcement Learning Without Wasting Rollouts on Hundred-Billion-Scale MoE