Code-LLM Vulnerability Recurrence Evaluation on FSTab LLM-generated software
51.23FVRGemini-3 Pro
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Gemini-3 Pro2026.02 | 51.23 | 25.09 | 41.39 | 58.67 | |
| Gemini-3 Flash2026.02 | 44.29 | 29.77 | 36.1 | 52.81 | |
| Composer2026.02 | 43.86 | 35.53 | 46.43 | 57.32 | |
| GPT-5.22026.02 | 37.52 | 23.2 | 33.92 | 42.3 | |
| Claude-4.5 Opus2026.02 | 35.37 | 21.44 | 31.75 | 53.58 | |
| Grok2026.02 | 31.43 | 11.96 | 27.85 | 57.29 |