Infrastructure Safety Assessment on Infrastructure Mutation Scenarios safe-unsafe trials (Held-out)
0.3Unsafe RateSQA
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SQA2026.06 | 0.3 | 3.1 | 93.9 | 99.7 | 96.7 | |
| SAV2026.06 | 18.5 | 8.7 | 81.8 | 81.5 | 81.6 | |
| HomoQ2026.06 | 21 | 6.8 | 84.8 | 79 | 81.8 | |
| Static OPA2026.06 | 34.7 | 2.1 | 93.3 | 65.3 | 76.8 |