← All workshopsRead Variational Autoencoders

Generative Models

Sample a code without losing the controls

BOTIE

Convert the supplied log variances to the scales that multiply standard-normal noise.
Ready to test
0 passed · 0 failed · 1 not run

An empty floor. Start with the information your machine needs.

Received result

Your answer will arrive here.

Examples, experiments and workspace controls
Live values

Connect a tool's inputs to see what it produces. You can try a branch before connecting it to Result.

Try your own inputs

Use a copy of the example to try your own data. Keep the problem's assumptions in mind, and predict the effect of a change before checking the output. You can also add another branch to compare two ways of doing the calculation.

Start at an output arrow, then click or drag to an input. Drag empty space to pan. Delete removes a selection. Ctrl/Cmd+Z undoes a change.
Context for this challenge

Convert the supplied log variances to the scales that multiply standard-normal noise.

We want to choose new latent codes from a distribution, rather than only reuse codes assigned to training examples. A model has supplied a mean and log variance. Let's combine them with a standard-normal draw to construct a sample with the intended location and spread.

Inspect the component calculation
Exponentiate(
  Scores: Multiply(
    Value: 0.5
    Multiplier: Latent log variance
  )
)
Examples and results
Latent log variance
  1. 00
  2. 10
Required result
  1. 01
  2. 11
Try a fresh construction

This clears this draft, including its undo history. Earlier successful attempts stay recorded. Build again without opening hints to record an independent attempt.

Stuck? Inspect a worked construction

The tools below form one possible solution. You can assemble it, follow its signals, and undo that change to return to your own construction.

Assumptions and test precision

This is a small construction problem. The lesson explains the assumptions and the parts of the full method that this example leaves out. Numerical tests allow a difference of 0.001. Lists and structures must also have the right shape.

Your last tested pieces can start the full construction. An existing draft there keeps its own pieces.

The latent coordinates have different spreads

This check uses the pieces from “Turn a standard draw into a latent sample”.

BOTIE

Botie's encoder supplies two latent means and log variances. The first coordinate has variance 4 and the second variance 0.25. Use the supplied standard-normal draw to produce one latent sample.
Latent mean
  1. 02
  2. 1-1
Latent log variance
  1. 01.3863
  2. 1-1.3863
Standard-normal draw
  1. 0-0.5
  2. 12

Use a number, text, or an array matching the requested result. For a matrix, put each row in brackets. Saving your prediction runs your current construction on these inputs and checks the earlier examples too.

Build a construction in the workspace above first. This case will test those pieces.

Which supplied values determine location, spread, and the particular sampled deviation?

Sampling a code is one component; a VAE objective must also shape its encoded distributions.

This is a small construction problem. The lesson explains the assumptions and the parts of the full method that this example leaves out.

Work through the Python challenges →
Continue the courseGenerative Adversarial Networks →
← Autoencoders