SemAttack: Natural Textual Attacks via Different Semantic Spaces
This addresses the vulnerability of language models to adversarial attacks, which is a security concern for AI applications, but it is incremental as it builds on existing attack methods by improving efficiency and naturalness.
The authors tackled the problem of generating natural adversarial text to attack pre-trained language models, proposing SemAttack, which achieved high attack success rates against state-of-the-art models like DeBERTa-v2 and defenses like FreeLB, with human evaluations confirming the naturalness of the generated text.
Recent studies show that pre-trained language models (LMs) are vulnerable to textual adversarial attacks. However, existing attack methods either suffer from low attack success rates or fail to search efficiently in the exponentially large perturbation space. We propose an efficient and effective framework SemAttack to generate natural adversarial text by constructing different semantic perturbation functions. In particular, SemAttack optimizes the generated perturbations constrained on generic semantic spaces, including typo space, knowledge space (e.g., WordNet), contextualized semantic space (e.g., the embedding space of BERT clusterings), or the combination of these spaces. Thus, the generated adversarial texts are more semantically close to the original inputs. Extensive experiments reveal that state-of-the-art (SOTA) large-scale LMs (e.g., DeBERTa-v2) and defense strategies (e.g., FreeLB) are still vulnerable to SemAttack. We further demonstrate that SemAttack is general and able to generate natural adversarial texts for different languages (e.g., English and Chinese) with high attack success rates. Human evaluations also confirm that our generated adversarial texts are natural and barely affect human performance. Our code is publicly available at https://github.com/AI-secure/SemAttack.