#exploration-exploitation tradeoff

Shanghai AI Lab researchers propose entropy-based scaling laws and novel techniques to overcome exploration collapse in reinforcement learning for reasoning-centric large language models, achieving significant performance improvements.