PaperGym
โ† Back to Foundations
๐Ÿง 
Module 5 of 6

Your First Neural Network

Tie everything together โ€” build a working neural network from scratch.

Progress0/3 drills ยท 0%
FREE~35 minยทBeginner-Intermediate

What you'll learn

  • โœ“Linear layer forward pass
  • โœ“Two-layer network with ReLU
  • โœ“Backpropagation and a training loop

Drills

1

Linear Layer (Forward)

Implement a linear layer forward pass โ€” the fundamental building block of every neural network.

Easy๐Ÿ• 8mโšก 10 pts
2

Two-Layer Network (Forward)

Chain two linear layers with ReLU to build a complete classifier โ€” input โ†’ hidden โ†’ output โ†’ softmax.

Medium๐Ÿ• 10mโšก 15 pts
3

Backpropagation & Training Loop

Compute gradients with the chain rule and train a network on XOR until convergence โ€” the capstone drill.

Hard๐Ÿ• 15mโšก 25 pts

๐Ÿ“š Resources