| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| TruthfulQA MC2 | MC2 Accuracy56.46 | 46 | 25d ago | ||
| TruthfulQA f_con o=5 | Acc (Exact)100 | 18 | 1mo ago | ||
| TruthfulQA | ϵ-FTRL | Mean Per-Step Regret0.138 | 15 | 1mo ago | |
| TruthfulQA | IPO | Accuracy (TruthfulQA)53.9 | 12 | 1mo ago | |
| TruthfulQA | Llama | Normalized Probability Mass33.7 | 3 | 1mo ago | |
| TruthfulQA | SEA | TruthRate83 | 3 | 1mo ago | |
| TruthfulQA o=5 (f_clean) | Accuracy (Exact)55.5 | 3 | 1mo ago |