Zero-shot Question Answering on ARC Easy
33.53AccuracyGELU
Evaluation results
| Method | Links | |
|---|---|---|
| GELUArchitecture=GELU MLP, Evaluation mode=zero-shot2026.05 | 33.53 | |
| Yat (pn+ca)Parameterization=per-neuron, Scaling=constant per-channel, Evaluation mode=zero-shot2026.05 | 33.02 | |
| Yat (pn+α)Parameterization=per-neuron, Scaling=learnable per-channel, Evaluation mode=zero-shot2026.05 | 33.01 | |
| Yat (sb+ca)Parameterization=shared (b, ε), Scaling=constant per-channel, Evaluation mode=zero-shot2026.05 | 32.98 | |
| Yat (sb+α)Parameterization=shared (b, ε), Scaling=learnable per-channel, Evaluation mode=zero-shot2026.05 | 32.62 |