neural_lam.models.latent.constant_encoder

Constant (input-independent) latent encoder.

Classes

ConstantLatentEncoder

Latent encoder that returns a constant (input-independent) distribution.

Module Contents

class neural_lam.models.latent.constant_encoder.ConstantLatentEncoder(latent_dim: int, num_mesh_nodes: int, output_dist: str = 'isotropic')

Bases: neural_lam.models.latent.base_encoder.BaseLatentEncoder

Latent encoder that returns a constant (input-independent) distribution.

compute_dist_params returns a tensor of zeros, so the resulting Normal is Normal(mean=0, std=1) for output_dist="isotropic" and Normal(mean=0, std=softplus(0)+eps) for output_dist="diagonal".

Store the number of mesh nodes to produce parameters for.

Parameters:

• latent_dim (int) – Dimensionality of the latent variable at each mesh node.

• num_mesh_nodes (int) – Number of mesh nodes the latent variable is defined on.

• output_dist (str) – Type of output distribution: "isotropic" or "diagonal".

compute_dist_params(grid_rep: torch.Tensor, **kwargs: Any) → torch.Tensor

Compute raw distribution parameters from the grid representation.

For this constant encoder the parameters are all zeros, independent of the values in grid_rep.

Parameters:

• grid_rep (torch.Tensor) – Shape (B, num_grid_nodes, d_h). Grid input representation, used only to determine batch size and device.

• **kwargs – Ignored; accepted for compatibility with the base class interface.

Returns:

Shape (B, num_mesh_nodes, output_dim). Raw parameters of the latent distribution, all zeros.

Return type:

torch.Tensor

num_mesh_nodes