omnibias.torch.activations
PyTorch activation derivatives, trainable operator layers, network jets, and direct differential-operator contractions.
import omnibias.torch.activationsModule contract & mathematical conventions
Activation dictionary: registry plus the activation families.
Importing this module triggers registration of every activation in
smooth, proximal, classical, nqs, trigonometric,
piecewise (hard non-smooth, almost-everywhere towers), and tempered
(smooth beta-tempered surrogates). Use get_activation or
list_activations to access them by name.
Shared formulas do not imply universal floating-point identity. Verify dtype, shape, derivative-order support, and the stated numerical assumptions for the path you use.
classicalModule ↗nqsModule ↗piecewiseModule ↗proximalModule ↗registryModule ↗smoothModule ↗temperedModule ↗trigonometricModule ↗Imports & aliases
These names share a definition with another public module. Follow a name for its full contract, parameters, and examples.
| Import from this module | Canonical reference |
|---|---|
ActivationSpec | omnibias.core.ActivationSpec |
get_activation | omnibias.torch.get_activation |
is_registered | omnibias.torch.is_registered |
list_activations | omnibias.torch.list_activations |
register_activation | omnibias.torch.register_activation |