nemos.glm.PopulationGLM.feature_mask

property PopulationGLM.feature_mask: Array | dict[str, Array]

Mask indicating which features are used for each neuron.

The feature mask has a tree structure matching the coefficients (coef_):

• Array input: Shape (n_features, n_neurons). Each entry [i, j] indicates whether feature i is used for neuron j (1 = used, 0 = masked).

• Dict/FeaturePytree input: A dict with keys matching coef_. Each leaf array has shape (n_neurons,), indicating whether that feature group is used for each neuron.

Returns:

The feature mask, or None if not set.

Return type:

jnp.ndarray or dict[str, jnp.ndarray]