Course Study Note

Day 3 — Variational Autoencoders

I studied why the encoder predicts a mean and variance, then connected reparameterization to reconstruction and KL losses.

Date
2025-07-09
Course
2025 Generative Models Intensive — From VAE to Stable Diffusion
01VAE Architecture and ELBO

Study checkpoint for this day

02Reparameterization Trick

Study checkpoint for this day

A day-level study path generated from the saved course topic order.

Topics from the day

  1. VAE Architecture and ELBO

    I diagrammed how a VAE learns a latent distribution rather than a single code.

  2. Reparameterization Trick

    I checked how sampling is rewritten so gradients can pass through the computation.

VAE encoder predicting mean and variance before sampling a latent code
The encoder predicts a latent distribution rather than a single coordinate.

From an autoencoder to a VAE

A VAE predicts the mean and variance of (q_\phi(z x)). I reread Auto-Encoding Variational Bayes to connect the approximate posterior, prior, and lower bound.

Reparameterization

z = mu + std * eps does not remove randomness. It moves the random variable outside the learned parameters so gradients can reach both mu and std.

Two loss terms

Reconstruction preserves information from the input; KL divergence discourages the posterior from drifting too far from the prior. Overweighting the KL term can lead the decoder to ignore the latent code.

The saved latent-interpolation output was unavailable, so I stopped at the architecture and loss path and will rerun that comparison separately.

← Course home
Publication first-page preview