| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| SCIFACT abstract-level (test) | Qwen-7B | Support F189.9 | 12 | 1mo ago | |
| Matter-of-Fact scientific claims | GPT-5-mini + Analytica-L | Overall Accuracy73 | 9 | 2mo ago | |
| SciFact | mistral-large | Accuracy40.5 | 6 | 4mo ago | |
| SciFact (test) | GPT-4 Turbo | Precision (NE)93 | 4 | 3mo ago | |
| NealSmith abstract-level (test) | DeepSeek-R1 | SUP F1 Score69.03 | 2 | 1mo ago |