RAP: Robustness-Aware Perturbations for Defending against Backdoor Attacks on NLP Models

TL;DR

This paper introduces RAP, an online defense method that uses robustness-aware perturbations to effectively detect and defend against backdoor attacks in NLP models, outperforming existing methods in accuracy and efficiency.

Contribution

The paper proposes a novel robustness-aware perturbation technique for backdoor defense in NLP, with theoretical analysis and superior experimental results.

Findings

Achieves better defense performance than existing methods.

Maintains lower computational costs.

Effectively distinguishes poisoned samples from clean ones.

Abstract

Backdoor attacks, which maliciously control a well-trained model's outputs of the instances with specific triggers, are recently shown to be serious threats to the safety of reusing deep neural networks (DNNs). In this work, we propose an efficient online defense mechanism based on robustness-aware perturbations. Specifically, by analyzing the backdoor training process, we point out that there exists a big gap of robustness between poisoned and clean samples. Motivated by this observation, we construct a word-based robustness-aware perturbation to distinguish poisoned samples from clean samples to defend against the backdoor attacks on natural language processing (NLP) models. Moreover, we give a theoretical analysis about the feasibility of our robustness-aware perturbation-based defense method. Experimental results on sentiment analysis and toxic detection tasks show that our method achieves better defending performance and much lower computational costs than existing online defense methods. Our code is available at https://github.com/lancopku/RAP.