测试大模型删知识方法，发现删不干净还易被绕过。
Does Unlearning Truly Unlearn? A Black Box Evaluation of LLM Unlearning Methods
- 用新生物数据集和WMDP测试删知识效果
- 删后性能下降明显，5次提示可使准确率提升10倍以上
- 学无关数据几乎能完全恢复原性能，说明没真正删除
大型语言模型（LLM）的删知识旨在移除有害信息以防止其被用于恶意用途。现有方法如LLMU和RMU在删知识基准上表现优异。本文通过WMDP数据集及自建生物学数据集，深入评估删知识对模型性能的影响。结果表明，删知识显著损害模型通用能力，且LLMU影响更严重。进一步测试显示，仅通过5次提示或简单改写问题，即可使删知识基准上的准确率提升超十倍。此外，训练无关数据几乎完全恢复删前性能，证明当前方法未能真正实现删知识。本研究提供了一套评估删知识方法的黑盒框架，代码已公开于https://github.com/JaiDoshi/Knowledge-Erasure。
原文摘要 · Abstract (English)
Large language model unlearning aims to remove harmful information that LLMs have learnt to prevent their use for malicious purposes. LLMU and RMU have been proposed as two methods for LLM unlearning, achieving impressive results on unlearning benchmarks. We study in detail the impact of unlearning on LLM performance metrics using the WMDP dataset as well as a new biology dataset we create. We show that unlearning has a notable impact on general model capabilities, with the performance degradation being more significant in general for LLMU. We further test the robustness of the two methods and find that doing 5-shot prompting or rephrasing the question in simple ways can lead to an over ten-fold increase in accuracy on unlearning benchmarks. Finally, we show that training on unrelated data can almost completely recover pre-unlearning performance, demonstrating that these methods fail at truly unlearning. Our methodology serves as an evaluation framework for LLM unlearning methods. The code is available at: https://github.com/JaiDoshi/Knowledge-Erasure.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。