Atomic Skills are the Prerequisite: When Reinforcement Learning Synthesizes Compositional Reasoning, and When It Only Amplifies 文章

ArXiv CS.CL2026-05-28NEWSen作者: Sitao Cheng, Xunjian Yin, Ruiwen Zhou, Yuxuan Li, Xinyi Wang, Liangming Pan, William Yang Wang, Victor Zhong

查看原文 →

Atomic Skills are the Prerequisite: When Reinforcement Learning Synthesizes Compositional Reasoning, and When It Only Amplifies · 相关技术

相关技术

Retrieval-augmented generation Supervised Fine-Tuning Contextual Reasoning Parametric Reasoning Continual Learning Complementary Reasoning reinforcement learning