Tier

Intermediate

Go from 'it trains' to understanding why — convolutions, generalization, and the knobs that control training itself.

1. 1Convolutions VisualizedSlide a kernel across an image by hand and watch the exact math tf.conv2d runs internally.~20 min
2. 2Overfitting & RegularizationShrink the training set, grow the model, and watch train/validation loss diverge — then fight back with L2 and dropout.~25 min
3. 3Training Dynamics LabCompare optimizers, learning rates, and batch sizes side by side and see why the same model trains completely differently.~25 min
4. 4Explainability BasicsA trained model's gradient with respect to its input is a tiny saliency map — drag a probe point to see what it's sensitive to, and why.~15 min
5. 5Word EmbeddingsTrain tiny 2D word vectors from scratch and watch related words cluster together purely from co-occurrence statistics.~20 min
6. 6Sequence Models: RNNsOrder-sensitive data needs a network with memory — step through a recurrent cell by hand, then train one to detect palindromes.~25 min