Multi-task Language Understanding on MMLU-pro
14,946.31Average Context Length (tokens)PRD
Evaluation Results
| Method | Links | |
|---|---|---|
| PRDType=Full Debate2025.07 | 14,946.31 | |
| PRDType=Full Debate2025.07 | 14,829.46 | |
| MLDType=Full Debate2025.07 | 11,412.75 | |
| RECONCILEType=Full Debate2025.07 | 11,334.12 | |
| RECONCILEType=Full Debate2025.07 | 11,260.15 | |
| MLDType=Full Debate2025.07 | 11,213.07 | |
| ChatEvalType=Full Debate2025.07 | 9,219.88 | |
| ChatEvalType=Full Debate2025.07 | 9,160.49 | |
| GDType=Part Debate2025.07 | 7,265.23 | |
| GDType=Part Debate2025.07 | 7,159.45 | |
| NDType=Part Debate2025.07 | 6,673.75 | |
| NDType=Part Debate2025.07 | 6,647.27 | |
| OursType=Part Debate2025.07 | 6,001.33 | |
| OursType=Part Debate2025.07 | 5,897.94 | |
| MaVType=No Debate2025.07 | 2,268.1 | |
| MaVType=No Debate2025.07 | 2,144.39 |