Secure Code Generation on LLMSecEval 150 tasks
0Number of VulnerabilitiesMA-CoT
Evaluation Results
| Method | Links | |
|---|---|---|
| MA-CoTLanguage=C, Model=gpt-52026.05 | 0 | |
| MA-CoTLanguage=C, Model=claude-4.52026.05 | 0 | |
| ZeroShotLanguage=Java, Model=gpt-52026.05 | 0 | |
| CoTLanguage=Java, Model=gpt-52026.05 | 0 | |
| MA-CoTLanguage=Java, Model=gpt-52026.05 | 0 | |
| ZeroShotLanguage=Python, Model=gpt-52026.05 | 0 | |
| MA-CoTLanguage=Python, Model=gpt-52026.05 | 0 | |
| MA-CoTLanguage=Python, Model=claude-4.52026.05 | 0 | |
| MA-CoTLanguage=Python, Model=gemini-2.52026.05 | 0 | |
| CoTLanguage=C, Model=gemini-2.52026.05 | 1 | |
| MA-CoTLanguage=C, Model=gemini-2.52026.05 | 1 | |
| MA-CoTLanguage=Java, Model=gemini-2.52026.05 | 1 | |
| VanillaLanguage=Python, Model=gpt-52026.05 | 1 | |
| CoTLanguage=Python, Model=gpt-52026.05 | 1 | |
| VanillaLanguage=Java, Model=gpt-52026.05 | 2 | |
| MA-CoTLanguage=Java, Model=claude-4.52026.05 | 2 | |
| ZeroShotLanguage=Python, Model=claude-4.52026.05 | 2 | |
| VanillaLanguage=C, Model=claude-4.52026.05 | 3 | |
| ZeroShotLanguage=C, Model=claude-4.52026.05 | 3 | |
| ZeroShotLanguage=Java, Model=gemini-2.52026.05 | 3 | |
| CoTLanguage=Java, Model=gemini-2.52026.05 | 3 | |
| CoTLanguage=Python, Model=claude-4.52026.05 | 3 | |
| CoTLanguage=Python, Model=gemini-2.52026.05 | 3 | |
| VanillaLanguage=C, Model=gpt-52026.05 | 4 | |
| VanillaLanguage=C, Model=gemini-2.52026.05 | 4 | |
| VanillaLanguage=Python, Model=claude-4.52026.05 | 4 | |
| VanillaLanguage=Python, Model=gemini-2.52026.05 | 4 | |
| ZeroShotLanguage=C, Model=gemini-2.52026.05 | 5 | |
| VanillaLanguage=Java, Model=gemini-2.52026.05 | 5 | |
| CoTLanguage=Java, Model=claude-4.52026.05 | 6 | |
| ZeroShotLanguage=Python, Model=gemini-2.52026.05 | 6 | |
| ZeroShotLanguage=C, Model=gpt-52026.05 | 7 | |
| CoTLanguage=C, Model=claude-4.52026.05 | 7 | |
| CoTLanguage=C, Model=gpt-52026.05 | 10 | |
| ZeroShotLanguage=Java, Model=claude-4.52026.05 | 28 | |
| VanillaLanguage=Java, Model=claude-4.52026.05 | 46 |