Forget to Know, Remember to Use: Context-Aware Unlearning for Large Language Models
This addresses a usability gap in unlearning for large language models, ensuring models can still leverage removed information contextually, which is incremental but important for practical deployment.
The paper tackles the problem that existing unlearning methods for large language models impair the model's ability to use removed knowledge when it is reintroduced in the prompt, and proposes a plug-in term that restores this contextual utility to near original levels while maintaining effective forgetting and retain-set utility.
Large language models may encode sensitive information or outdated knowledge that needs to be removed, to ensure responsible and compliant model responses. Unlearning has emerged as an efficient alternative to full retraining, aiming to remove specific knowledge while preserving overall model utility. Existing evaluations of unlearning methods focus on (1) the extent of forgetting of the target knowledge (forget set) and (2) maintaining performance on the retain set (i.e., utility). However, these evaluations overlook an important usability aspect: users may still want the model to leverage the removed information if it is re-introduced in the prompt. In a systematic evaluation of six state-of-the-art unlearning methods, we find that they consistently impair such contextual utility. To address this, we augment unlearning objectives with a plug-in term that preserves the model's ability to use forgotten knowledge when it is present in context. Extensive experiments demonstrate that our approach restores contextual utility to near original levels while still maintaining effective forgetting and retain-set utility.