Beyond Trajectory Imitation: Strategy-Guided Policy Optimization for LLM Reasoning
cs.AI
Submitted: 2026-06-23
Updated: 2026-10-13
Code: https://github.com/huggingface/math-verify
Terminology
Sources
- Step 3.5 Flash: Open Frontier-Level Intelligence with 11B Active Parameters
- SFT or RL? An Early Investigation into Training R1-Like Reasoning Large Vision-Language Models
- SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training
- The Llama 3 Herd of Models
- Measuring Mathematical Problem Solving With the MATH Dataset
- UFT: Unifying Supervised and Reinforcement Fine-Tuning
- Towards a Unified View of Large Language Model Post-Training
- Learning What Reinforcement Learning Can't: Interleaved Online Fine-Tuning for Hardest Questions
- Overcoming Exploration in Reinforcement Learning with Demonstrations
- OpenAI o1 System Card
- Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Qwen2.5 Technical Report
- LIMO: Less is More for Reasoning
- DAPO: An Open-Source LLM Reinforcement Learning System at Scale
- More Than One Teacher: Adaptive Multi-Guidance Policy Optimization for Diverse Exploration
- BREAD: Branched Rollouts from Expert Anchors Bridge SFT & RL for Reasoning
- Proximal Supervised Fine-Tuning
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems