| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| QAC Evaluation Dataset 2026 preprint (offline) | Coverage0.9772 | 8 | 3mo ago | ||
| DocQAC SU | Global-Guided BART | MRR0.711 | 6 | 1mo ago | |
| Human Evaluation Set | Full | Item-wise Score69.9 | 4 | 3mo ago | |
| DocQAC SS | DocQ-Guided BART | MRR0.72 | 3 | 1mo ago | |
| Live production traffic (10%) | Avg Characters Typed0 | 2 | 3mo ago |
