← Back to Architecture
manifold-guided attention
ArchitectureUsed in
1 PRs
Best BPB
0.4380
Avg BPB
0.4380
Submissions
Hyperparameters Across PRs
| pr_number | parameters |
|---|---|
| 663 | {"heads":2} |
| pr_number | parameters |
|---|---|
| 663 | {"heads":2} |