Model Registry · FieldTransformer family
FieldTransformer v1.4
MultiScaleFieldLayer (k=3 + k=7 Laplacian) + causal attention · flux / Schwarzian / latent-diffusion / entropy regularizers
STATUS TRAININGPROJECT: FIELDTRANSFORMER
Benchmarks
| Metric | Value | Notes |
|---|---|---|
| v1.2 final loss | 0.0109 | PPL 1.01 — memorization at this scale, labeled as such |
| v1.3c | KILLED | stopped at 3,500 steps |
Known Limitations
- −Tiny-corpus results only; PPL 1.01 on Shakespeare demonstrates fit, not generalization.
- −Custom IP terms apply (see repository notice).