Counterfactual Explanations for Models of Code

Cito, Jürgen; Dillig, Işıl; Murali, Vijayaraghavan; Chandra, Satish

doi:10.1109/icse-seip55303.2022.9794112

Cited by 8 publications

(3 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…What's worse, since the existing posthoc explanation approaches mainly leverage perturbation-based mechanisms (e.g., LEMNA [30]) to track input features that are highly relevant to the model's prediction, the explanation performance will deteriorate further due to the weak robustness of detection models to random perturbations. On the contrary, counterfactual explanations [18] contain the most crucial information, which constitutes minimal changes to the input under which the model changes its mind. However, just because of this, they may only cover a small subset of the ground truth.…”

Section: Why Not Existing Explainers?mentioning

confidence: 99%

Coca: Improving and Explaining Graph Neural Network-Based Vulnerability Detection Systems

Cao,

Sun,

et al. 2024

Proceedings of the IEEE/ACM 46th International Conference on Software Engineering

View full text Add to dashboard Cite

Recently, Graph Neural Network (GNN)-based vulnerability detection systems have achieved remarkable success. However, the lack of explainability poses a critical challenge to deploy blackbox models in security-related domains. For this reason, several approaches have been proposed to explain the decision logic of the detection model by providing a set of crucial statements positively contributing to its predictions. Unfortunately, due to the weaklyrobust detection models and suboptimal explanation strategy, they have the danger of revealing spurious correlations and redundancy issue.In this paper, we propose Coca, a general framework aiming to 1) enhance the robustness of existing GNN-based vulnerability detection models to avoid spurious explanations; and 2) provide both concise and effective explanations to reason about the detected vulnerabilities. Coca consists of two core parts referred to as Trainer and Explainer. The former aims to train a detection model which is robust to random perturbation based on combinatorial contrastive learning, while the latter builds an explainer to derive crucial code statements that are most decisive to the detected vulnerability via dual-view causal inference as explanations. We

show abstract

Section: Why Not Existing Explainers?mentioning

confidence: 99%

Coca: Improving and Explaining Graph Neural Network-Based Vulnerability Detection Systems

Cao,

Sun,

et al. 2024

Proceedings of the IEEE/ACM 46th International Conference on Software Engineering

View full text Add to dashboard Cite

show abstract

“…Besides the attention mechanism, various explainable AI approaches have also been employed to explain NMT models of source code. For example, Cito et al [13] integrated counterfactual explanation techniques for NMT models that predict certain properties of code or code changes. Rabin et al [36] provided a model-agnostic approach to identify critical input features for models of code and demonstrated that the approach enables code simplification in code search and variable misuse debugging.…”

Section: Explainable Nmt-based Code Generationmentioning

confidence: 99%

On the Reliability and Explainability of Automated Code Generation Approaches

Liu¹,

Tantithamthavorn²,

Liu³

et al. 2023

Preprint

View full text Add to dashboard Cite

Automatic code generation, the task of generating new code snippets from existing code or comments, has long been of interest. Numerous code generation models have been proposed and proven on different benchmark datasets. However, little is known about whether this objective has been achieved and why code generation models effectively transform code sequences automatically. In other words, can we totally trust these automated code generation models? Consequently, there is a pressing need to understand the inner logic of code generation models and to investigate their replicability, reliability, and explainability. To bridge these research gaps, we conduct a thorough empirical study of five code generation models on four representative code generation datasets to assess the limits and capabilities of automatic code generation approaches. We further employ advanced explainable AI approaches to highlight the tokens that significantly contribute to the code generation. Experiments demonstrate that we successfully replicate state-of-the-art code generation approaches. We discover that state-of-the-art approaches suffer from severe data duplication and input insensitivity, which are subtle issues with significant implications. Our explainability analysis reveals that, in various experimental scenarios, code generation models can recognize code grammar and structural information, but can not capture key tokens that need to be updated. Our results draw several lessons and guidelines for future work in this area.CCS Concepts: • Software and its engineering → Software maintenance tools; • General and reference → Reliability; • Computing methodologies → Natural language processing.

show abstract

“…Meanwhile, the security issues of these models have also become a growing concern. Recent studies [7], [8], [9], [10], [11], [12] reveal that many language models of code [13], [14], [15], [16] (a.k.a., code models) can produce opposite results for two inputs that share the same program semantics, one of which is generated by applying semanticpreserving transformations (e.g., variable renaming) to the other.…”

Section: Introductionmentioning

confidence: 99%

Stealthy Backdoor Attack for Code Models

Zhang¹,

Xu²,

Zhang³

et al. 2023

Preprint

View full text Add to dashboard Cite

Code models, such as CodeBERT and CodeT5, offer general-purpose representations of code and play a vital role in supporting downstream automated software engineering tasks. Most recently, code models were revealed to be vulnerable to backdoor attacks. A code model that is backdoor-attacked can behave normally on clean examples but will produce pre-defined malicious outputs on examples injected with triggers that activate the backdoors. Existing backdoor attacks on code models use unstealthy and easy-to-detect triggers. This paper aims to investigate the vulnerability of code models with stealthy backdoor attacks. To this end, we propose AFRAIDOOR (Adversarial F eatur e as Adaptive Backdoor ). AFRAIDOOR achieves stealthiness by leveraging adversarial perturbations to inject adaptive triggers into different inputs. We evaluate AFRAIDOOR on three widely adopted code models (CodeBERT, PLBART and CodeT5) and two downstream tasks (code summarization and method name prediction). We find that around 85% of adaptive triggers in AFRAIDOOR bypass the detection in the defense process. By contrast, only less than 12% of the triggers from previous work bypass the defense. When the defense method is not applied, both AFRAIDOOR and baselines have almost perfect attack success rates. However, once a defense is applied, the success rates of baselines decrease dramatically to 10.47% and 12.06%, while the success rate of AFRAIDOOR are 77.05% and 92.98% on the two tasks. Our finding exposes security weaknesses in code models under stealthy backdoor attacks and shows that the state-of-the-art defense method cannot provide sufficient protection. We call for more research efforts in understanding security threats to code models and developing more effective countermeasures.

show abstract

Counterfactual Explanations for Models of Code

Cited by 8 publications

References 33 publications

Coca: Improving and Explaining Graph Neural Network-Based Vulnerability Detection Systems

Coca: Improving and Explaining Graph Neural Network-Based Vulnerability Detection Systems

On the Reliability and Explainability of Automated Code Generation Approaches

Stealthy Backdoor Attack for Code Models

Contact Info

Product

Resources

About