Machine unlearning aims to forget sensitive knowledge from Large Language Models (LLMs) while maintaining general utility. However, existing approaches typically treat all tokens in a response indiscriminately and enforce uncertainty over the entire vocabulary. This global treatment results in unnecessary utility degradation and extends optimization to content-agnostic regions. To address these limitations, we propose PALU (Prefix-Aware Localized Unlearning), a framework driven by a local entropy maximization objective across both temporal and vocabulary dimensions. PALU reveals that (i) suppressing the sensitive prefix alone is sufficient to sever the causal generation link, and (ii) flattening only the top-$k$ logits is adequate to maximize uncertainty in the critical subspace. These findings allow PALU to avoid redundant optimization across the full vocabulary and parameter space while minimizing collateral damage to general model performance. Extensive experiments validate that PALU achieves superior forgetting efficacy and utility preservation compared to state-of-the-art baselines.

翻译：机器遗忘旨在使大语言模型（LLMs）遗忘敏感知识，同时保持其通用效用。然而，现有方法通常不加区分地处理响应中的所有词元，并在整个词汇表上强制施加不确定性。这种全局处理方式会导致不必要的效用下降，并将优化扩展至与内容无关的区域。为应对这些局限性，我们提出了PALU（前缀感知局部化遗忘），这是一个由跨时间和词汇维度的局部熵最大化目标驱动的框架。PALU揭示：（i）仅抑制敏感前缀便足以切断因果生成链；（ii）仅展平前$k$个逻辑值便足以在关键子空间内最大化不确定性。这些发现使得PALU能够避免在整个词汇表和参数空间上进行冗余优化，同时将对模型通用性能的附带损害降至最低。大量实验验证，与最先进的基线方法相比，PALU在遗忘效能和效用保持方面均实现了更优的性能。