KnowRL: Boosting LLM Reasoning via Reinforcement Learning with Minimal-Sufficient Knowledge Guidance

Benchmark Model Rank Results
mathematical-reasoning-on-hmmt-feb-2025KnowRL-Nemotron-1.5B (CSS)#15Accuracy: 48.75
mathematical-reasoning-on-aime24KnowRL-Nemotron-1.5B (CBRS)#41Acc: 75.52
mathematical-reasoning-on-aime-2025KnowRL-Nemotron-1.5B (CSS)#63Accuracy: 65.21