Differentiable Language Model Adversarial Attacks on Categorical Sequence Classifiers

Abstract: An adversarial attack paradigm explores various scenarios for the vulnerability of deep learning models: minor changes of the input can force a model failure. Most of the state of the art frameworks focus on adversarial attacks for images and other structured model inputs, but not for categorical sequences models. Successful attacks on classifiers of categorical sequences are challenging because the model input is tokens from finite sets, so a classifier score is non-differentiable with respect to inputs, and gradient-based attacks are not applicable. Common approaches deal with this problem working at a token level, while the discrete optimization problem at hand requires a lot of resources to solve. We instead use a fine-tuning of a LLM for adversarial attacks as a generator of adversarial examples. To optimize the model, we define a differentiable loss function that depends on a surrogate classifier score and on a deep learning model that evaluates approximate edit distance. So, we control both the adversability of a generated sequence and its similarity to the initial sequence. As a result, we obtain semantically better samples. Moreover, they are resistant to adversarial training and adversarial detectors. Our model works for diverse datasets on bank transactions, electronic health records, and NLP datasets.