🕸️ Dynamic Computation Graph Visualizer

Watch PyTorch build its graph live, one line at a time — then watch gradients flow backward through it for real.

📋 How To Use This Lab
  1. Press Next Step ▶ to execute the next line of code — a new node or edge appears as it runs.
  2. Use Previous Step ◀ to rewind, or Auto Run ▶▶ to watch the whole script play itself.
  3. The last line, c.backward(), sends real gradient values flowing backward through every edge.
  4. Watch x especially — it feeds two different paths, so its gradient is the sum of both.
Grapher
🤖 GRAPHERPress Next Step to run the first line — I'll draw every node and edge as PyTorch creates them, live.
🧑‍🏫 PROF. TORCHThis is a DYNAMIC graph — it's built fresh on every forward pass, one operation at a time, not compiled ahead of time.