Question Answering Accuracy on ARC C
70.14Accuracy (ARC-C)MERGEvolve
Evaluation Results
| Method | Links | |
|---|---|---|
| MERGEvolve2026.06 | 70.14 | |
| Pack of LLMsSetting=single-task, Number of runs=52026.06 | 70.14 | |
| LoraHubSetting=single-task, Number of runs=52026.06 | 69.97 | |
| MERGEvolveSetting=single-task, Number of runs=52026.06 | 69.65 | |
| EMMSetting=single-task, Number of runs=52026.06 | 69.45 | |
| Model Swarms2026.06 | 69.09 | |
| Pack of LLMs2026.06 | 69.03 | |
| EMM2026.06 | 69.03 | |
| Expert FusionSetting=single-task, Number of runs=52026.06 | 69.03 | |
| LoraHub2026.06 | 68.87 | |
| Model SwarmsSetting=single-task, Number of runs=52026.06 | 68.6 | |
| Expert Fusion2026.06 | 68.08 | |
| Best Single ExpertSetting=single-task, Number of runs=52026.06 | 57.34 | |
| NITPModel Scale=3B2026.05 | 51.55 | |
| NTPModel Scale=3B2026.05 | 51.2 | |
| Data MergeSetting=single-task, Number of runs=52026.06 | 46.08 | |
| NITPModel Scale=2B2026.05 | 41.92 | |
| NTPModel Scale=2B2026.05 | 39.52 | |
| NITPModel Scale=0.5B2026.05 | 34.71 | |
| NTPModel Scale=0.5B2026.05 | 31.95 | |
| TIESSetting=single-task, Number of runs=52026.06 | 31 |