| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Counterfactual Reasoning | CRAFT Hard Split (test) | Accuracy83.64 | 8 | |
| Counterfactual Reasoning | CRAFT Easy Split (test) | Accuracy80.05 | 8 | |
| Physical Reasoning Visual Question Answering | CRAFT VQA | Descriptive Accuracy39.4 | 4 | |
| Named Entity Recognition | CRAFT | F1 Score0.7577 | 4 |