MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement

arXiv:2610.11959 · cs.CL · Submitted 2026-10-08 · Read on arXiv

cs.CL

Submitted: 2026-10-08

Updated: 2026-10-08

Code: https://github.com/datacurve-ai/deep-swe

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Benchmark — 26× in this paper · explained in 2 episodes
• Reward Hacking — 12× in this paper · explained in 8 episodes
• Reinforcement Learning — 11× in this paper · explained in 4 episodes
• Distillation — 10× in this paper · explained in 2 episodes
• Consistency — 9× in this paper · explained in 3 episodes
• Core — 9× in this paper · explained in 2 episodes
• GRPO — 8× in this paper · explained in 6 episodes
• Groupwise Advantage Redistribution — 8× in this paper
• Verification — 8× in this paper · explained in 2 episodes
• Verifier — 8× in this paper · explained in 2 episodes
• Actor — 7× in this paper · explained in 3 episodes
• Alignment — 7× in this paper · explained in 1 episode
• Transformer — 6× in this paper · explained in 2 episodes
• Aggregation — 5× in this paper · explained in 1 episode
• Fidelity — 5× in this paper · explained in 2 episodes
• Groupwise Reward Synthesis — 5× in this paper
• AI Agents — 4× in this paper · explained in 2 episodes
• Accuracy — 4× in this paper · explained in 1 episode
• Embeddings — 4× in this paper · explained in 2 episodes
• Global Attention — 4× in this paper
• Robustness — 4× in this paper · explained in 5 episodes
• Speculative Decoding — 4× in this paper · explained in 2 episodes
• Adam — 3× in this paper · explained in 2 episodes
• Agentic Task — 3× in this paper · explained in 1 episode
• Agentic Tasks — 3× in this paper · explained in 1 episode
• Benchmarking — 3× in this paper · explained in 3 episodes
• Credit assignment — 3× in this paper · explained in 3 episodes
• Fine-Tuning — 3× in this paper · explained in 5 episodes
• Foundation Model — 3× in this paper · explained in 3 episodes
• Generalization — 3× in this paper · explained in 13 episodes
• On-Policy Distillation — 3× in this paper · explained in 3 episodes
• Block Diffusion — 2× in this paper · explained in 2 episodes
• Decode — 2× in this paper · explained in 2 episodes
• Feed-Forward Network — 2× in this paper
• Foundation Models — 2× in this paper · explained in 7 episodes
• Quantization — 2× in this paper · explained in 12 episodes
• Reinforcement Learning (RL) — 2× in this paper · explained in 25 episodes
• Tokenization — 2× in this paper · explained in 2 episodes
• Trace — 2× in this paper · explained in 2 episodes
• AdamW Optimizer — 1× in this paper · explained in 1 episode
• Agentic Coding — 1× in this paper · explained in 1 episode
• Agentic Evaluation — 1× in this paper · explained in 1 episode
• Automatic Speech Recognition (ASR) — 1× in this paper · explained in 2 episodes
• Contrastive Learning — 1× in this paper · explained in 5 episodes
• Grounding — 1× in this paper · explained in 3 episodes
• Group Relative Policy Optimization — 1× in this paper
• Group Relative Policy Optimization (GRPO) — 1× in this paper · explained in 5 episodes
• KV cache — 1× in this paper · explained in 3 episodes
• Knowledge Distillation — 1× in this paper · explained in 11 episodes
• Local Sliding Window Attention — 1× in this paper
• Mixture-of-Experts (MoE) — 1× in this paper · explained in 7 episodes
• Model Context Protocol — 1× in this paper
• Model Context Protocol (MCP) — 1× in this paper · explained in 8 episodes
• Multi-hop Reasoning — 1× in this paper · explained in 3 episodes
• Multimodal Inputs — 1× in this paper · explained in 2 episodes
• Quantization-Aware Training — 1× in this paper · explained in 3 episodes
• Regularization — 1× in this paper · explained in 3 episodes
• Representation Learning — 1× in this paper · explained in 2 episodes
• Supervised Fine-Tuning (SFT) — 1× in this paper · explained in 8 episodes
• Teacher-Student Gap — 1× in this paper · explained in 2 episodes

Sources

• SWE-rebench V2: Language-Agnostic SWE Task Collection at Scale
• DFlash: Block Diffusion for Flash Speculative Decoding
• MiMo-V2-Flash Technical Report
• SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?
• AsyncFlow: An Asynchronous Streaming RL Framework for Efficient LLM Post-Training
• Semantic Head Specialization Guides Hybrid ViT Attention for Multimodal LLMs
• DeepSWE: Measuring Frontier Coding Agents on Original, Long-Horizon Engineering Tasks
• SEC-bench Pro: Can Language Models Solve Long-Horizon Software Security Tasks?
• ExploitBench: A Capability Ladder Benchmark for LLM Cybersecurity Agents
• The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution
• JobBench: Aligning Agent Work With Human Will
• Multi-Turn On-Policy Distillation with Prefix Replay
• Muown: Row-Norm Control for Muon Optimization
• DeepSeek-V3 Technical Report
• Muon is Scalable for LLM Training
• Stabilizing MoE Reinforcement Learning by Aligning Training and Inference Routers
• MOPD: Multi-Teacher On-Policy Distillation for Capability Integration in LLM Post-Training
• Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces
• OfficeQA Pro: An Enterprise Benchmark for End-to-End Grounded Reasoning
• GDPval: Evaluating AI Model Performance on Real-World Economically Valuable Tasks

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering