bound_lower_power

bound_lower_power(param: Tensor, update: Tensor, limit: float, *, power: float, **kwargs) → Tensor

Computes the scaled update of lower-bound power parameter dependence.

This is sometimes also referred to as “soft parameter dependence”.

\[U_- = (P - P_\text{min})^{\mu_-} U_-\]

Parameters:

• param (torch.Tensor) – parameter with update bounding, \(P\).

• update (torch.Tensor) – depressive update being applied, \(U_-\).

• limit (float) – value of the upper bound, \(P_\text{min}\).

• power (float) – exponent of parameter dependence, \(\mu_-\).

Returns:

bounded update.

Return type:

torch.Tensor