| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Deceptive Alignment Benchmark (DAB) 400 scenarios | MechELK | Elicitation Accuracy81.2 | 12 | 5d ago | |
| Quirky LM 1,200 factual questions | MechELK | Elicitation Accuracy0.874 | 12 | 5d ago | |
| TruthfulQA MC1 | MechELK | Elicitation Accuracy86.7 | 12 | 5d ago |