Atlas

Foundations

Recurrent Networks

Memory over time, and why it eventually broke.

You can jump straight in, but this star assumes Training & Optimization. Not recommended as a first stop.

Four layers of depth

Each layer ends with a quiz. Finish layer 4 and you own this concept.

  1. L1IntuitionWhy sequences (text, audio, time series) need networks with memory, and what an RNN intuitively does.5m
    40 XP
  2. L2MechanicsThe recurrence equation, why vanilla RNNs struggle with long sequences, and how LSTMs/GRUs fix it with gates.8m
    70 XP
  3. L3CodeImplementing a vanilla RNN cell from scratch and an LSTM-based sequence model in PyTorch.12m
    110 XP
  4. L4FoundationsBackpropagation through time, why vanishing gradients are provably worse for recurrence, and the LSTM's gating math.15m
    180 XP

33 stars in the atlas.