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

ArXi:2512.01970v3 Announce Type: replace Does Reinforcement Learning (RL) merely amplify existing skills, or synthesize novel skills? We investigate this question through the lens of Complementary Reasoning: the critical practical capability of integrating internal knowledge with external context, a prerequisite for reliable Continual Learning and Retrieval-Augmented Generation. To avoid pre-