| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| CVE-based Vulnerability Dataset | Success Rate (Rs)10 | 29 | 3mo ago | ||
| CyberGym (evaluation) | GPT-5.4 mini Best-of-8 | Total Rollouts1,600 | 15 | 23d ago | |
| OSS-Fuzz 100 random issues | ARVO | Successful Reproduction Count81 | 10 | 1mo ago | |
| ReDoS CVEs | PUFFERDOS | Exploit Rate96.8 | 3 | 1mo ago | |
| FL-Bench 100 (standardized) | PoC-Adapt | Success Rate (SR)15 | 2 | 3mo ago | |
| n-day vulnerabilities (reproduction set) | - | - | 0 | 5mo ago |
