AIFuniverseAI很好玩
← Back to Play Now
Play Now

GAN Lab: Train Two Adversarial Networks by Hand

GAN Lab

Choose or draw a two-dimensional data distribution, start generator-versus-discriminator training, and watch fake samples, decision boundaries, and gradients change in real time.

Time to tryTry it in 15 seconds
CostFree

Why it’s interesting

Choose or draw a two-dimensional data distribution, start generator-versus-discriminator training, and watch fake samples, decision boundaries, and gradients change in real time.

An abstract GAN becomes a visible chase. Draw an unusual shape and watch the generator struggle to bend random points into a matching distribution.

How to try it

Open the page, choose a point pattern or draw one, then press play. You can also change the learning rate or step through training to inspect each update.

Use the official page or demonstration as the current reference. Do not assume that a staged installation, limited experiment, account-gated demo, subscription feature, or physical product is available everywhere.

Requirements and sources

This entry credits Minsuk Kahng、Nikhil Thorat、Duen Horng ‘Polo’ Chau、Fernanda Viégas、Martin Wattenberg and keeps “GAN Lab” as the original title. It is free to view or try and available as an immediate browser or media experience. Pricing, access, regional availability, accounts, hardware, and service support can change, so verify them on the [first-party source](https://poloclub.github.io/ganlab/) before trying or buying. The description stays within the documented AI role and does not treat promotional claims as independent testing.

Was this project worth the click?

Anonymous feedback helps us improve the selection. We do not build visitor profiles.