[2603.01285] Attention Smoothing Is All You Need For Unlearning

Computer Science > Machine Learning arXiv:2603.01285 (cs) [Submitted on 1 Mar 2026] Title:Attention Smoothing Is All You Need For Unlearning Authors:Saleh Zare Zade, Xiangyu Zhou, Sijia Liu, Dongxiao Zhu View a PDF of the paper titled Attention Smoothing Is All You Need For Unlearning, by Saleh Zare Zade and 3 other authors View PDF HTML (experimental) Abstract:Large Language Models are prone to memorizing sensitive, copyrighted, or hazardous content, posing significant privacy and legal concerns. Retraining from scratch is computationally infeasible, whereas current unlearning methods exhibit unstable trade-offs between forgetting and utility, frequently producing incoherent outputs on forget prompts and failing to generalize due to the persistence of lexical-level and semantic-level associations in attention. We propose Attention Smoothing Unlearning (ASU), a principled framework that casts unlearning as self-distillation from a forget-teacher derived from the model's own attention. By increasing the softmax temperature, ASU flattens attention distributions and directly suppresses the lexical-level and semantic-level associations responsible for reconstructing memorized knowledge. This results in a bounded optimization objective that erases factual information yet maintains coherence in responses to forget prompts. Empirical evaluation on TOFU, MUSE, and WMDP, along with real-world and continual unlearning scenarios across question answering and text completion, demonstr...