Autoencoder — Encoder-Decoder & Latent Space
Compress high-dim data to a bottleneck, then reconstruct. Interpolate in latent space.
Controls
Latent dim:
2
Noise level:
0.10
Interpolation α:
0.50
New point A
New point B
Re-encode
Sweep interpolation
Encoder:
Input → latent z (bottleneck)
Decoder:
z → reconstruction x̂
Loss:
||x - x̂||²
Interpolation:
z_α = (1-α)·z_A + α·z_B
x̂_α = Decoder(z_α)
Latent space is smooth — linear interpolation gives meaningful outputs.