A Black-Box Attack on Code Models via Representation Nearest Neighbor Search

TL;DR

This paper introduces RNNS, a black-box attack method on code models that uses representation nearest neighbor search to generate effective adversarial examples with minimal perturbations across multiple programming languages and models.

Contribution

RNNS leverages a search seed and vector space mapping to improve adversarial code example generation, reducing perturbations and verification costs compared to existing methods.

Findings

RNNS achieves higher attack success rates and query efficiency.

Adversarial examples have smaller perturbations than baseline methods.

RNNS effectively attacks defended models and aids adversarial training.

Abstract

Existing methods for generating adversarial code examples face several challenges: limted availability of substitute variables, high verification costs for these substitutes, and the creation of adversarial samples with noticeable perturbations. To address these concerns, our proposed approach, RNNS, uses a search seed based on historical attacks to find potential adversarial substitutes. Rather than directly using the discrete substitutes, they are mapped to a continuous vector space using a pre-trained variable name encoder. Based on the vector representation, RNNS predicts and selects better substitutes for attacks. We evaluated the performance of RNNS across six coding tasks encompassing three programming languages: Java, Python, and C. We employed three pre-trained code models (CodeBERT, GraphCodeBERT, and CodeT5) that resulted in a cumulative of 18 victim models. The results demonstrate that RNNS outperforms baselines in terms of ASR and QT. Furthermore, the perturbation of adversarial examples introduced by RNNS is smaller compared to the baselines in terms of the number of replaced variables and the change in variable length. Lastly, our experiments indicate that RNNS is efficient in attacking defended models and can be employed for adversarial training.