Erhan Xu

About

Erhan's research lies at the intersection of reinforcement learning and large language models. He works on statistically principled methods for LLM alignment and post-training of reasoning models. His broader interests include large-scale optimisation, cognitive science, agentic training, and causal inference/econometrics. Erhan welcomes collaboration with researchers in both academia and industry.