| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Machine-Generated Text Attribution | STEM | Full Macro F191.9 | 12 | |
| STEM tasks | STEM GPQA and MMLU redux (test) | GPQA Score30.3 | 12 | |
| STEM Problem Solving | STEM S-GPQA, GPQA-D, GSM8k, Math | S-GPQA25.81 | 12 | |
| Multiple-Choice Question Answering | STEM 4-Choice | Delta Accuracy0.3 | 6 |