| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| LLM Judge | ActAdd | Controllability Score90.38 | 40 | 1mo ago | |
| PolyGuard | ActAdd | PolyGuard100 | 40 | 1mo ago | |
| HarmBench | AAS | HarmBench Score87.5 | 40 | 1mo ago | |
| HDPCG Space Scale L (aggregated runs) | PF-A* | MAE (Density)1.343 | 3 | 1mo ago | |
| HDPCG Space Scale M (aggregated runs) | NP-A* | MAE (Density)1.366 | 3 | 1mo ago | |
| HDPCG Space Scale S (aggregated runs) | NP-A* | MAE (Density)2.217 | 3 | 1mo ago | |
| Harmless-helpful Group 4 (unseen) | MOC | Kendall's Tau1 | 2 | 12d ago | |
| Harmless-helpful Group 3 (unseen) | MOC | Kendall's Tau1 | 2 | 12d ago | |
| Harmless-helpful Group 2 (unseen) | MOC | Kendall's tau1 | 2 | 12d ago | |
| Harmless-helpful Group 1 (unseen) | MOC | Kendall's Tau1 | 2 | 12d ago | |
| Humor-helpful Group 4 (unseen) | MOC | Kendall's tau1 | 2 | 12d ago | |
| Humor-helpful Group 3 (unseen) | MOC | Kendall's tau1 | 2 | 12d ago | |
| Humor-helpful Group 2 (unseen) | MOC | Kendall's Tau1 | 2 | 12d ago | |
| Humor-helpful Group 1 (unseen) | MOC | Kendall's Tau1 | 2 | 12d ago |