# On the Origin and Evolution of Training-Inference Mismatch in LLM Reinforcement Learning

> Source: <https://aiflash.com/news/143897/>
> Published: 2026-11-02 04:30:00+00:00

On-policy reinforcement learning for large language models assumes that training responses are sampled from the current policy. In practice, an inference engine samples the responses and a separate trainer computes the gradients, and the training-inference mismatch between the two makes the training