Results for "stochastic regularization"

Stochastic generation strategies that trade determinism for diversity; key knobs include temperature and nucleus sampling.

A scalar measure optimized during training, typically expected loss over data, sometimes with regularization terms.

Autoencoder using probabilistic latent variables and KL regularization.

Measure of vector magnitude; used in regularization and optimization.

A gradient method using random minibatches for efficient training on large datasets.

Optimization under uncertainty.

Techniques that discourage overly complex solutions to improve generalization (reduce overfitting).