| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| PhyX mini (test) | GRPO (Fmt + Acc + ASM) | Overall Accuracy46.2 | 7 | 3d ago | |
| NegBench COCO subset | G-DINO-B + NEGTOME | Overall Accuracy32.55 | 4 | 25d ago | |
| CICERO v2 | DIALECT-Large | Macro F188.63 | 2 | 1mo ago | |
| CICERO | DIALECT-Large | Macro F170.66 | 2 | 1mo ago |