| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Consequential-Action Attack Success Rate | Pooled Cross-Model Evaluation Suite (all split) | Attack Success Rate (ASR)31.2 | 5 | |
| Consequential-Action Attack Success Rate | Pooled Cross-Model Evaluation Suite whitebox | ASR22.2 | 5 | |
| Consequential-Action Attack Success Rate | Pooled Cross-Model Evaluation Suite (adaptive split) | Attack Success Rate (ASR)27.6 | 5 | |
| Consequential-Action Attack Success Rate | Pooled Cross-Model Evaluation Suite static split | ASR (%)43.8 | 5 | |
| Utility evaluation | Pooled Cross-Model Evaluation Suite | Utility95.9 | 4 |