Attacks against Abstractive Text Summarization Models through Lead Bias and Influence Functions
This addresses security vulnerabilities in summarization models, which is an incremental but important domain-specific problem for AI safety and NLP practitioners.
The authors tackled the adversarial robustness of abstractive text summarization models by exploiting lead bias for perturbations and using influence functions for data poisoning, resulting in models producing extractive rather than abstractive summaries under attack.
Large Language Models have introduced novel opportunities for text comprehension and generation. Yet, they are vulnerable to adversarial perturbations and data poisoning attacks, particularly in tasks like text classification and translation. However, the adversarial robustness of abstractive text summarization models remains less explored. In this work, we unveil a novel approach by exploiting the inherent lead bias in summarization models, to perform adversarial perturbations. Furthermore, we introduce an innovative application of influence functions, to execute data poisoning, which compromises the model's integrity. This approach not only shows a skew in the models behavior to produce desired outcomes but also shows a new behavioral change, where models under attack tend to generate extractive summaries rather than abstractive summaries.