Xiaoyang Li

I’m a second year undergraduate student in University of California, Irvine.

Today with the advent of LLM, there are predictions of LLM becoming AGI. Although LLM achieve outstanding results on language related task, they still inevitably exhibit hallucination, catatrophic forgetting, etc. To solve these issues, my current research interest lies in continual learning and multimodal in LLM to mimic the way human learns and make model robust to variation and hallucinations.