Vulnerability Detection on Python Vulnerability Detection
99.04F1 Scorecodegen-mono
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| codegen-monoTechnique Used=Instruction Following Fine Tuning, Number of Parameters=350m parameters model2025.04 | 99.04 | 99 | 98.08 | 100 | — | |
| Farasat & PoseggaTechnique Used=ML (BiLSTM)2025.04 | 94.7 | 98.6 | 96.2 | 93.3 | 99.3 | |
| Dozono et al. (GPT-4o)Technique Used=LLM Prompting (Various strategies), Backbone / Model=GPT-4o2025.04 | 0.8 | — | — | — | — | |
| Dozono et al. (GPT-4T)Technique Used=LLM Prompting (Various strategies), Backbone / Model=GPT-4T2025.04 | 0.76 | — | — | — | — | |
| Dozono et al. (Gemini 1.5 Pro)Technique Used=LLM Prompting (Various strategies), Backbone / Model=Gemini 1.5 Pro2025.04 | 0.75 | — | — | — | — | |
| Dozono et al. (CodeLlama-7b)Technique Used=LLM Prompting (Various strategies), Backbone / Model=CodeLlama-7b2025.04 | 0.72 | — | — | — | — | |
| Dozono et al. (GPT-3.5T)Technique Used=LLM Prompting (Various strategies), Backbone / Model=GPT-3.5T2025.04 | 0.7 | — | — | — | — | |
| Dozono et al. (CodeLlama-13b)Technique Used=LLM Prompting (Various strategies), Backbone / Model=CodeLlama-13b2025.04 | 0.35 | — | — | — | — |