| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| CRAG | Truthfulness42.3 | 36 | 1mo ago | ||
| GPQA Diamond | GPQA Score69.19 | 16 | 1mo ago | ||
| OpenScience | P-POTS+Mirror | Pass@153.8 | 16 | 2mo ago | |
| GPQA | DIVER | Accuracy48.5 | 14 | 2mo ago | |
| GPQA | FailFast | Speedup3.11 | 12 | 4mo ago | |
| NQ | Truthfulness39.8 | 8 | 1mo ago | ||
| TriviaQA | Baseline | Pass@1 Accuracy71.7 | 8 | 1mo ago | |
| HotpotQA | Truthfulness63.6 | 6 | 1mo ago | ||
| HotpotQA (dev) | PMSR | Accuracy59.8 | 6 | 3mo ago | |
| Average (CRAG, NQ, HotpotQA, MuSiQue) | Truthfulness36.8 | 4 | 1mo ago | ||
| MuSiQue | Truthfulness2.9 | 4 | 1mo ago | ||
| MMLU Redux | MASRouter | Training Cost (USD)0.46 | 4 | 4mo ago | |
| MuSiQue | TruthRL | Truthfulness Score (T)-0.9 | 2 | 1mo ago |