ExploitGym kernel intended-vulnerability exploits
Agents · 2026-05-11
Count of ExploitGym Linux-kernel tasks exploited through the intended vulnerability with standard mitigations disabled. The official suite contains 186 kernel privilege-escalation tasks.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-5.6 Sol | 73.0 |
| GPT-5.5 | 21.0 |
| Claude Mythos Preview | 12.0 |
| Opus 4.6 | 1.0 |
| GPT-5.4 | 1.0 |
| Opus 4.7 | 0.0 |
| Gemini 3.1 Pro Preview | 0.0 |
| GLM-5.1 | 0.0 |
| Muse Spark 1.1 | 0.0 |