| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| CORD official (test) | UDoc | F1 Score98.94 | 37 | 4mo ago | |
| BioNLP | UNIVERSALNER | F1 Macro80.96 | 31 | 3mo ago | |
| MIT Restaurant | GPT-4 | F1-macro85.35 | 24 | 4mo ago | |
| BioNLP2004, MIT Movie, MIT Restaurant, and OntoNotes 5.0 (Average) few-shot (test) | NuNER | Avg F1 (few-shot)71.53 | 21 | 4mo ago | |
| SCIERC (test) | SpERT | F1 Score70.33 | 20 | 4mo ago | |
| Mol-Instructions | Cuttlefish | F1 Score78 | 13 | 4mo ago | |
| Public Thyroid Ultrasound Datasets OOD (test) | EchoVLM | Hamming Accuracy62.4 | 12 | 3mo ago | |
| Public Liver Ultrasound Datasets OOD (test) | EchoVLM | Hamming Accuracy91.06 | 12 | 3mo ago | |
| Public Breast Ultrasound Datasets OOD (test) | HuatuoGPT-Vision | Hamming Accuracy57.48 | 12 | 3mo ago | |
| In-house ER | Average Score17.52 | 11 | 16d ago | ||
| Aggregate Domain Medical and Banking | Qwen Omni-3-30BA3B | EER6.29 | 8 | 22d ago | |
| Banking Domain | EER8.88 | 8 | 22d ago | ||
| Medical Domain 8 held-out sets | Qwen Omni-3-30BA3B | EER5.84 | 8 | 22d ago | |
| ADE 10-fold cross-validation | Relation-Metric Network | F1 Score0.8711 | 6 | 4mo ago | |
| SCIERC (dev) | SCIIE | Precision70 | 6 | 4mo ago |
