Benchmark
CyberGym Level 0 success rate
Higher scores are better.
Benchmark sourceScores from CyberGym leaderboard data docs.
- Models measured
- 1
- Scores recorded
- 1
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 0.035
- Mean
- 0.035
- Median
- 0
- Std dev
- 0.035
- 10th percentile
- 0.035
- 90th percentile
- 0.0345
- Min
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | GPT 4.1agent: OpenHands, trials: unstated, date: 2025-05-15· CyberGym leaderboard data docs | 0.0345 |