从条件数视角提出模型免疫化新方法，提升对有害微调的鲁棒性。
Model Immunization from a Condition Number Perspective
- 基于海森矩阵条件数构建免疫化分析框架
- 通过正则化控制预训练后条件数，实现免疫化
- 在线性与非线性模型上均验证有效
模型免疫化旨在预训练出难以被有害任务微调但保持其他任务性能的模型。尽管已有研究在文本到图像模型中提供了免疫化的实证支持，但免疫化是否可行及免疫模型的准确定义仍不明确。本文提出一个基于海森矩阵条件数的分析框架，用于线性模型的免疫化分析，并据此设计一种包含正则化项的算法，以控制预训练后的条件数。在线性模型和非线性深度网络上的实验结果表明，该算法在实现模型免疫化方面具有有效性。代码已公开于 https://github.com/amberyzheng/model-immunization-cond-num。
原文摘要 · Abstract (English)
Model immunization aims to pre-train models that are difficult to fine-tune on harmful tasks while retaining their utility on other non-harmful tasks. Though prior work has shown empirical evidence for immunizing text-to-image models, the key understanding of when immunization is possible and a precise definition of an immunized model remain unclear. In this work, we propose a framework, based on the condition number of a Hessian matrix, to analyze model immunization for linear models. Building on this framework, we design an algorithm with regularization terms to control the resulting condition numbers after pre-training. Empirical results on linear models and non-linear deep-nets demonstrate the effectiveness of the proposed algorithm on model immunization. The code is available at https://github.com/amberyzheng/model-immunization-cond-num.
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。