Overcoming Scaling Limits in On-Policy Self-Distillation for LLM Reasoning

Open paper
Benchmark Model Rank Results
mathematical-reasoning-on-aime-2025OASIS (Qwen3-8B) (extra data)–Accuracy: 72.22
mathematical-reasoning-on-aime24OASIS (Qwen3-8B) (extra data)–Acc: 77.5

Rank counts only results with a code link.