| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| GPQA-Diamond | Base | Final Accuracy99.3 | 28 | 5mo ago | |
| SCIMDR Eval | Accuracy49.9 | 8 | 4mo ago | ||
| SPIQA | SPIQA-A79.9 | 8 | 4mo ago | ||
| CharXiv | CharXiv-D Score95.2 | 8 | 4mo ago | ||
| MedQA | Accuracy83.74 | 7 | 1mo ago | ||
| ChartQA | Qwen-3-VL-8B | Accuracy87.4 | 6 | 4mo ago | |
| Molecule captioning | MENTOR46 | 5 | 17d ago | ||
| Molecular description generation | SciReasoner | ROUGE-L75 | 5 | 17d ago | |
| Name conversion i2f | SciReasoner | Element Match93 | 5 | 17d ago | |
| Name conversion i2s | SciReasoner | Exact Match87 | 5 | 17d ago | |
| Name conversion s2f | SciReasoner | Element Match96 | 5 | 17d ago | |
| Name conversion s2i | SciReasoner | Split Match50 | 5 | 17d ago | |
| Open question | SciReasoner | BertScore0.85 | 5 | 17d ago | |
| Multiple choice question | Accuracy91 | 5 | 17d ago | ||
| Chemical disease interaction extraction | SciReasoner | F1 Score54 | 5 | 17d ago | |
| Chemical protein interaction extraction | SciReasoner | F1 Score36 | 5 | 17d ago | |
| Chemical entity recognition | SciReasoner | F1 Score88 | 5 | 17d ago | |
| GPQA | Teacher | Average Accuracy41.94 | 4 | 1mo ago |