Rewarding Graph Reasoning Process makes LLMs more Generalized Reasoners

Open in new window