news

Nov 15, 2025 Our paper on adversarial robustness of concept representations in sparse autoencoders was accepted to EACL 2026!
Jan 22, 2025 Our paper “More RLHF, More Trust?” was accepted to ICLR 2025 as an oral presentation (Top 1.8%)!
Aug 26, 2024 I am excited to start my PhD in Computer Science at UC Berkeley, advised by Prof. Bin Yu and Prof. Ion Stoica!
Jul 01, 2024 Our paper “Certifying LLM Safety against Adversarial Prompting” was accepted to COLM 2024!
May 01, 2024 Our paper “Improving Prototypical Visual Explanations with R3” was accepted to ICML 2024!