Sort on Sort OOD lengths: 2x, 4x, 8x
100Exact Match AccuracyASEntmax
Evaluation Results
| Method | Links | |
|---|---|---|
| ASEntmaxLayers (L)=2, OOD Sequence Length Scale=2x2025.06 | 100 | |
| EntmaxLayers (L)=2, OOD Sequence Length Scale=2x2025.06 | 99.3 | |
| Top-K (K=32)Layers (L)=2, OOD Sequence Length Scale=2x2025.06 | 92.5 | |
| ASEntmaxLayers (L)=2, OOD Sequence Length Scale=4x2025.06 | 79.7 | |
| EntmaxLayers (L)=2, OOD Sequence Length Scale=4x2025.06 | 57.8 | |
| SoftmaxLayers (L)=2, OOD Sequence Length Scale=2x2025.06 | 0 | |
| SoftmaxLayers (L)=2, OOD Sequence Length Scale=4x2025.06 | 0 | |
| SoftmaxLayers (L)=2, OOD Sequence Length Scale=8x2025.06 | 0 | |
| Top-K (K=32)Layers (L)=2, OOD Sequence Length Scale=4x2025.06 | 0 | |
| Top-K (K=32)Layers (L)=2, OOD Sequence Length Scale=8x2025.06 | 0 | |
| SSMaxLayers (L)=2, OOD Sequence Length Scale=2x2025.06 | 0 | |
| SSMaxLayers (L)=2, OOD Sequence Length Scale=4x2025.06 | 0 | |
| SSMaxLayers (L)=2, OOD Sequence Length Scale=8x2025.06 | 0 | |
| EntmaxLayers (L)=2, OOD Sequence Length Scale=8x2025.06 | 0 | |
| ASEntmaxLayers (L)=2, OOD Sequence Length Scale=8x2025.06 | 0 |