Tag
NudgeRL is a framework that enhances reinforcement learning with verifiable rewards (RLVR) by introducing structured exploration and strategy nudging, achieving better reasoning performance in large language models more efficiently than brute-force scaling methods.