Reasoning on DROP (test)
35.13AccuracyMoE w/ optimal AR
Evaluation results
| Method | Links | |
|---|---|---|
| MoE w/ optimal ARModel Type=MoE, Activation rate=20.07, Compute=2.86e21, Data reuse=-2025.06 | 35.13 | |
| Dense baselineModel Type=Dense, Compute=5.45e212025.06 | 32.32 | |
| MoE w/ optimal ARModel Type=MoE, Activation rate=20.07, Compute=2.86e21, Data reuse=strict2025.06 | 30.93 |