← Back to CCAB Gallery

Neural Network Playgrounds

Interactive visualizations exploring the fundamentals of deep learning, from simple perceptrons to transformers and beyond.

20
Interactive Demos
ML
Machine Learning

Fundamentals

01

Perceptron Learning

Step-by-step visualization of how a single-layer perceptron learns to classify data by adjusting weights.

Classification Interactive
02

Gradient Descent Landscape

Explore 3D loss surfaces and watch different optimizers (SGD, Momentum, Adam) navigate to the minimum.

3D Optimization
03

Activation Function Explorer

Compare ReLU, Sigmoid, Tanh, GELU, Swish and more with interactive graphs showing functions and derivatives.

Comparison Educational
04

Backpropagation Flow

Watch gradients propagate backward through a neural network during training with animated flow visualization.

Animation Training

Deep Learning Architectures

05

CNN Filter Viewer

Apply convolutional filters to images and see edge detection, blur, sharpen, and emboss effects in real-time.

CNN Convolution
06

Attention Mechanism

Visualize self-attention weights in transformers, showing how each token attends to others in a sequence.

Transformers NLP

Clustering & Unsupervised Learning

07

K-Means Clustering

Step through the K-means algorithm watching centroids dance as they converge to cluster centers.

Clustering Step-by-step

Coming Soon

08

Vanishing Gradient Visualizer

See how gradients shrink in deep networks and understand why this causes training problems.

Coming Soon
09

GAN Training Dynamics

Watch generator and discriminator compete in an adversarial game to create realistic samples.

Coming Soon
10

Autoencoder Latent Space

Explore the compressed representation learned by autoencoders and generate new samples.

Coming Soon