提出可精确删除微调数据的模型融合方法，避免信息泄露风险。
Exact Unlearning of Finetuning Data via Model Merging at Scale
- 通过局部掩码和全局符号向量约束，实现高效精准去训练。
- 在500个模型融合场景中，准确率提升5-80%，计算量减少250倍。
- 适合需要严格数据删除的AI系统，如合规性要求高的大模型应用。
近似去训练因效率高而受到欢迎，但实际中易被攻击导致已删除信息泄露。为解决此问题，本文提出SIFT-Masks（SIgn-Fixed Tuning-Masks），一种基于模型融合的精确去训练方法。该方法克服标准模型融合的两大缺陷：（1）融合大量任务会严重损害性能；（2）提升性能的方法通过跨任务共享信息，使精确去训练成本过高。SIFT-Masks通过（1）引入局部掩码恢复任务特定性能；（2）将微调约束于全局符号向量，实现合并前独立确定掩码。在最多融合500个模型的四种设置下，相较于简单融合，精度提升5-80%，且精确去训练所需计算量比其他融合基线低至250倍。
原文摘要 · Abstract (English)
Approximate unlearning has gained popularity as an approach to efficiently update an LLM so that it behaves (roughly) as if it was not trained on a subset of data to begin with. However, existing methods are brittle in practice and can easily be attacked to reveal supposedly unlearned information. To alleviate issues with approximate unlearning, we instead propose SIFT-Masks (SIgn-Fixed Tuning-Masks), an exact unlearning method based on model merging. SIFT-Masks addresses two key limitations of standard model merging: (1) merging a large number of tasks can severely harm utility; and (2) methods that boost utility by sharing extra information across tasks make exact unlearning prohibitively expensive. SIFT-Masks solves these issues by (1) applying local masks to recover task-specific performance; and (2) constraining finetuning to align with a global sign vector as a lightweight approach to determine masks independently before merging. Across four settings where we merge up to 500 models, SIFT-Masks improves accuracy by 5-80% over naive merging and uses up to 250x less compute for exact unlearning compared to other merging baselines.
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。