CyberGym-E2E Benchmark
Explore score history ↗CyberGym-E2E evaluates AI agents on end-to-end cybersecurity tasks, covering vulnerability discovery, proof-of-concept generation, and patch generation, using 920 real-world vulnerabilities from 139 open-source projects.
- Released
- 2026-06-03
- Institutions
- University of California, Berkeley · Johns Hopkins University
- Evaluates
- Cybersecurity, Language & Knowledge, cs.CR
- Openness
- unknown
- Importer
- Claire Radar
- Review status
- unreviewed
CyberGym-E2E paper
No reported scores are on record for this benchmark yet.