Multiple Choice Question Answering on ARC Challenge o.o.d.
90.65AccuracyWALLA I
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| WALLA Ipooling=log-pool2026.07 | 90.65 | 2.96 | 79.94 | 30.41 | 9.16 | |
| RouteLLM2026.07 | 90.61 | 2.37 | 89.18 | 46.38 | 9.35 | |
| Gemma22026.07 | 90.43 | 7.55 | 89.18 | 38.43 | 10.67 | |
| StackedGen2026.07 | 90.35 | 2.19 | 89.04 | 46.05 | 8.85 | |
| WALLA II2026.07 | 90.28 | 6.17 | 88.07 | 42.23 | 10.59 | |
| NIRTRouter2026.07 | 90.17 | 5.12 | 89.18 | 46.38 | 10.65 | |
| WALLA I2026.07 | 89.89 | 6.24 | 86.76 | 44.9 | 11.06 | |
| WALLA Icalibration=calibrated2026.07 | 89.59 | 8.64 | 78.77 | 50.24 | 9.84 | |
| SelfCertainty2026.07 | 89.22 | 2.7 | 91.59 | 66.96 | 10.71 | |
| RouterDC2026.07 | 87.48 | 7.22 | 63.21 | 27.5 | 13.98 | |
| UniformAvg2026.07 | 87.39 | 7.4 | — | 0 | 14.23 | |
| PackLLM2026.07 | 86.26 | 6.87 | 36.38 | -21.8 | 14.93 | |
| Llama3.12026.07 | 81.3 | 3.06 | 37.43 | -12.2 | 21.01 | |
| Aloe2026.07 | 80.35 | 8.58 | 51.17 | -0.43 | 23.43 | |
| BioMistral2026.07 | 66.61 | 10.95 | 30.68 | -25.8 | 40.65 |