Circuit Capstone

The capstone runs the whole course loop on one five-qubit circuit. Build it. Work out its width, gate counts and depth by hand. Write down a prediction with numbers. Run it with a fixed seed. Then break exactly one step and find the first wrong state. A circuit is finished when someone else can repeat that full record, not when the histogram just looks right.

What does the capstone ask you to do?

Everything in this course fits into one loop. You run it on one five-qubit circuit you can share:

Build a five-qubit GHZ circuit (from Building GHZ States).
Analyse it before running: width, gate count, two-qubit count, depth, and two-qubit depth (from Circuit Depth).
Predict the measurement result with numbers, in writing, before you run it.
Run it with a fixed seed and a fixed shot count. A seed is the starting number for the random draws, so the same seed gives the same samples every time. Then compare with your prediction.
Break one thing on purpose. Then find the first wrong state, like a debugger would (from Debugging Quantum Circuits).
Report the whole record so someone else can repeat it.

The order is on purpose. A prediction written after you see the results can never prove your thinking wrong. A prediction written before can. And being proved wrong is where the learning happens.

It is like a science fair project. You write your guess before the experiment, not after. Unlike most science fair projects, this one gives an exact answer to check against, because the simulator has no noise.