Next-token prediction on OpenWebText (train)
18.4PPL
Hermite activation
Updated 4d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2025.02 | 18.4 | 2.9122 | |
2025.02 | 18.85 | 2.9365 | |
2025.02 | 18.95 | 2.9418 | |
2025.02 | 19.05 | 2.9471 | |
2025.02 | 19.46 | 2.9684 |
Hermite activation
| Method | Links | ||
|---|---|---|---|
2025.02 | 18.4 | 2.9122 | |
2025.02 | 18.85 | 2.9365 | |
2025.02 | 18.95 | 2.9418 | |
2025.02 | 19.05 | 2.9471 | |
2025.02 | 19.46 | 2.9684 |