| Waypoint-OPD: Transferring RL-Induced Policy Shifts via Gated On-Policy Distillation | ✓ Link | 96.09 | Waypoint-OPD (Qwen3-14B) | 2026-10-15 |
| TGRL: Temperature-Grouped Reinforcement Learning for Efficient Exploration in LLMs | ✓ Link | 92.7 | TGRL (Qwen3-14B) (extra data) | 2026-09-27 |
| Understanding On-Policy Distillation: A Mechanistic Interpretability Perspective via Sparse Crosscoders | ✓ Link | 86.9 | OPD student (JustRL) | 2026-09-28 |
| | 82.5 | QWQ-32B-preview | |
| LoRA: Low-Rank Adaptation of Large Language Models | ✓ Link | 82 | Math-Master | 2021-06-17 |
| On the Off-Policy Teacher in On-Policy Distillation | | 81.1 | SCOUT (Skywork-OR1-Math-7B → DeepSeek-R1-Distill-Qwen-1.5B) | 2026-09-29 |
| ARISE: Agent Reasoning with Intrinsic Skill Evolution in Hierarchical Reinforcement Learning | ✓ Link | 75.4 | ARISE + GRPO (Qwen3-4B-Instruct-2507) | 2026-03-17 |
| Recovering Reliable Supervision via Critique-Guided On-Policy Distillation | ✓ Link | 74.7 | CriticOPD (Olmo-3-7B-Instruct-SFT) | 2026-10-14 |
| Parameter Exploration for RLVR via Variational Learning | ✓ Link | 73.44 | M3PO (Qwen2.5-Math-7B) | 2026-08-10 |
| Understanding Evolution Strategies for LLM Reasoning: Broader Reasoning Coverage than GRPO | | 72.3 | DeepSeek-R1-Distill-Qwen-1.5B (ES to GRPO) | 2026-08-27 |
| KV4-RL: Unlocking Faster Asynchronous Rollouts for RLVR with INT4 KV Cache Quantization | | 68.9 | KV4-RL (Qwen3-8B-Base) | 2026-10-10 |
| Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvement | | 62.5 | Qwen2.5-Math-7B-instruct | 2024-09-18 |
| Rethinking RL for LLM Reasoning: It's Sparse Policy Selection, Not Capability Learning | ✓ Link | 48.4 | ReasonMaxxer (Qwen2.5-Math-7B) | 2026-05-07 |
| OPMD: On-Policy Mechanism Distillation Beyond Output Supervision | | 40.16 | OPMD (RL-Math -> Qwen3-1.7B-Base) | 2026-09-29 |
| SkillFactory: Self-Distillation For Learning Cognitive Behaviors | ✓ Link | 37.5 | Qwen2.5-7B-Instruct (SkillFactory with 1k SFT + GRPO) | 2025-12-03 |