PhD student at the OPTML Group, Michigan State University, advised by Prof. Sijia Liu. Research scientist intern at the MIT-IBM Computing Research Lab.

Trustworthy Machine Learning. Machine Unlearning, Alignment & RLHF, Adversarial Machine Learning
Generative AI. Large-language Model, Multi-modal Language Model, Reasoning, Post-training, On-policy Distillation

Papers, news, and photographs.

LLM-Biometrics · tracing LLM lineage from weight geometry
Unlearn-Backdoor · attention sinks as a gateway for backdooring unlearning
Split_Adaptation · adapting a pre-trained vision transformer without exposing data or weights