Research map / Adversarial robustness and security / Attacks on text and graphs
Find Me if You Can: Deep Software Clone Detection by Exploiting the Contest between the Plagiarist and the Detector
Yan-Ya Zhang, Ming Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| train detector adversarially with plagiarist-generated clones instead of self-training on its own labels (ablation) | ACD detector · BigCloneBench | F1: 0.76 -> 0.92 | Improved |
| train detector adversarially with plagiarist-generated clones instead of self-training on its own labels (ablation) | ACD detector · OJClone | F1: 0.19 -> 0.73 | Improved |
| adversarial plagiarist training instead of supervised CDLH | ACD · BigCloneBench | F1: 0.82 -> 0.92 | Improved |
| adversarial plagiarist training instead of supervised CDLH | ACD · OJClone | F1: 0.57 -> 0.73 | Improved |
| adversarial training improves recall over CDLH | ACD · BigCloneBench | recall: 0.74 -> 0.88 | Improved |
| ACD vs lexical/syntactic detector SourcererCC | ACD · BigCloneBench | recall: 0.02 -> 0.88 | Improved |
| ACD vs syntax-based Deckard | ACD · OJClone | precision: 0.99 -> 0.73 | Hurt |
| ACD vs Deckard | ACD · OJClone | recall: 0.05 -> 0.73 | Improved |
| generated structural clones applied to CDLH-style detector trained on original data | CDLH · BigCloneBench generated clones | clone pairs detected: 459 of 1145 missed | Hurt |
| vary batch size of generated clones from 16 to 512 | ACD · BigCloneBench | F1 | No clear effect |
| vary batch size of generated clones from 16 to 512 | ACD · OJClone | F1 | Improved |
| vary number of applied changes from 1 to 8 | ACD · BigCloneBench and OJClone | F1 | No clear effect |
Closest papers by meaning
- AdaCCD: Adaptive Semantic Contrasts Discovery Based Cross Lingual Adaptation for Code Clone Detection (AAAI 2024 · similarity 0.83)
- Detecting AI-Generated Code Assignments Using Perplexity of Large Language Models (AAAI 2024 · similarity 0.79)
- Uncovering LLM-Generated Code: A Zero-Shot Synthetic Code Detector via Code Rewriting (AAAI 2025 · similarity 0.79)
- ChatGPT-Generated Code Assignment Detection Using Perplexity of Large Language Models (Student Abstract) (AAAI 2024 · similarity 0.78)
- Exploring Unbiased Deepfake Detection via Token-Level Shuffling and Mixing (AAAI 2025 · similarity 0.78)
- Is Function Similarity Over-Engineered? Building a Benchmark (NeurIPS 2024 · similarity 0.78)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
