一句话：提出基于可解释性的 INNER，通过神经元路由定位故障神经元并优化，实现安全公平鲁棒的神经网络修复。
深度神经网络面临后门等安全威胁，现有修复方法虽能修复缺陷，却可能产生类别准确率不平衡的偏见模型，或削弱对抗鲁棒性。作者提出 INNER，一种基于可解释性的神经网络修复方法。INNER 引入神经元路由，利用 model probe 评估各神经元对网络非期望行为的贡献，识别故障神经元并加以优化。在后门攻击、对抗攻击和错误预测三类场景上的实验表明，INNER 在保持准确率的同时兼顾公平性与鲁棒性；复用其发现的故障神经元还能提升其他修复方法的性能，体现实用性与通用性。
Along with the prevalent use of deep neural networks (DNNs), concerns have been raised on the security threats from DNNs such as backdoors in the network. While neural network repair methods have shown to be effective for fixing the defects in DNNs, they have been also found to produce biased models, with imbalanced accuracy across different classes, or weakened adversarial robustness, allowing malicious attackers to trick the model by adding small perturbations. To address these challenges, we propose INNER, an INterpretability-based NEural Repair approach. INNER formulates the idea of neuron routing for identifying fault neurons, in which the interpretability technique model probe is used to evaluate each neuron's contribution to the undesired behaviour of the neural network. INNER then optimizes the identified neurons for repairing the neural network. We test INNER on three typical application scenarios, including backdoor attacks, adversarial attacks, and wrong predictions. Our experimental results demonstrate that INNER can effectively repair neural networks, by ensuring accuracy, fairness, and robustness. Moreover, the performance of other repair methods can be also improved by re-using the fault neurons found by INNER, justifying the generality of the proposed approach.
DOI 原文 ·
@inproceedings{paperbot8605,
title = {Interpretability Based Neural Network Repair},
author = {Zuohui Chen and Jun Zhou and Youcheng Sun and Jingyi Wang and Qi Xuan and Xiaoniu Yang},
booktitle = {ISSTA},
year = {2024},
doi = {10.1145/3650212.3680330}
}