cs.AINov 7, 2026

Hallucination-Guided Unlearning: Using Hallucination Traces to Reveal Overfitted Memories

Authors: Jiayu Zhang, Ziqi Zhong, Yuliang Gai, Canran Xiao

Organizations: Peking University · UTS Business School, University of Technology Sydney · Massachusetts Institute of Technology · University of Massachusetts Boston · 

Abstract

Selective unlearning is increasingly important for the safety and compliance of large language models. Harmful or memorized content is often entangled with useful knowledge and difficult to enumerate. We use the model's own failures to identify candidates for unlearning. We propose Hallucination-Guided Unlearning (HGU), which uses hallucinations as diagnostic signals. HGU localizes unsupported spans via evidence and internal consensus, traces a sparse set of high-responsibility MLP neurons, and edits them with compact gates under a utility-preserving KL constraint. On TOFU, HGU achieves the best MU-FE trade-off, with Avg scores of 0.7221 (forget05) and 0.7699 (forget10), and the highest SepS H-Avg (0.4382/0.4715) under mixed forget/retain prompts. On the WMDP auxiliary task, HGU yields the strongest forgetting (e.g., 14.2 on Economics) while improving Retain (52.7) and MMLU (57.2). These results suggest that tracing hallucinations enables targeted, post-hoc unlearning that better balances forgetting efficacy and general utility without full retraining.

Figures & tables

Appendix figures & tables18 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. RA2^2: Retain-Anchored Attraction Preserves Forget-Adjacent Utility in LLM Unlearning

    Oct 14, 2026Zinan Ling, Yang Xiao, Ruimeng Ye +2Machine UnlearningLLM Unlearning

  2. Forgetting Only What Matters: Layer-Selective Unlearning toward Robust LLMs

    Sep 9, 2026Ravi Ranjan, Olivera Kotevska, Agoritsa PolyzouMemorization in Language ModelsLanguage Model Robustness

  3. Anatomy of Unlearning: The Dual Impact of Fact Salience and Model Fine-Tuning

    Feb 23, 2026Anna Borisiuk, Andrey Savchenko, Alexander Panchenko +1Machine UnlearningLLM Unlearning