API

Latent probe classes

ArithmeticProbe

A probe for exploring arithmetic operations in the latent space of a language model variational autoencoder (LM-VAE).

ClusterVisualizationProbe

A probe for visualizing the latent space of a language VAE via clustering techniques.

DisentanglementProbe

A probe for disentanglement metrics on the latent space of a language VAE.

InterpolationProbe

A probe for evaluating interpolation in the latent space of an LM-VAE.

TraversalProbe

A probe for analyzing latent space traversal in an LM-VAE.

Metrics

DisentanglementMetric

Metrics used to evaluate disentangled representations in latent variable models.

InterpolationMetric

Metrics used to evaluate interpolation in generative models and latent space representations.

Operations

ArithmeticOps

Vector arithmetic operations

InterpolationOps

Operations for obtaining and evaluating interpolations from a source (start) and target (end) representation vectors.

TraversalOps

Class for performing latent space traversal and distance computations.

Utility

select_models

Selects a list of LM-VAE models available from the neuro-symbolic-ai repository, according to the specified criteria.