Yihong Chen

PhD student, University College London

1 paper at NeurIPS 2025

Homepage· OpenReview· Semantic Scholar· Google Scholar

Poster Session 4

Thursday, December 4, 2025 · 4:30 PM → 7:30 PM

LLM Unlearning via Neural Activation Redirection

#1414 · William F. Shen, Xinchi Qiu, Meghdad Kurmanji, Alex Iacob, Lorenzo Sani, Yihong Chen, Nicola Cancedda, Nicholas D. Lane

We propose LUNAR, an novel LLM unlearning method by redirecting the representation of unlearned data for effective, controlled and efficient unlearning.