Cyber-security Vulnerability Discovery on CyberExplorer
90PrecisionClaude Opus 4.5
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Claude Opus 4.52026.02 | 90 | — | — | — | 22.5 | 47.3 | 4.52 | 173 | 127.4 | |
| Gemini 3 Pro2026.02 | 81.82 | — | — | — | 22.5 | 72.23 | 0.66 | 95 | 234.67 | |
| DeepSeek V32026.02 | 62.5 | — | — | — | 12.5 | 562.78 | 1.23 | 87 | 600.66 | |
| GPT 5.22026.02 | 60 | — | — | — | 15 | 101 | 1.4 | 105 | 231.76 | |
| Qwen 32026.02 | 17.65 | — | — | — | 7.5 | 674.95 | 1.23 | 122 | 208.08 |