Rethinking Training-Inference Mismatch in LLM Reinforcement Learning: Where It Arises and How to Correct It.

arXiv:2609.32444v1 Announce Type: cross Abstract: We study training-inference mismatch in reinforcement learning with verifiable rewards (RLVR) for large language models, where rollouts are sampled by an inference engine while gradients are computed by a train