Tutorial: Subjecting codes of various dimensionality to global rotations

[rafaelha]

Write a demo notebook that reproduces Fig. 4a of https://www.nature.com/articles/s41586-025-09848-5 using Tsim.

---

Goal

Add a tutorial notebook that reproduces Fig. 4a of the linked Nature paper — subjecting QEC codes of various dimensionality to global rotations — using tsim.

Deliverable

• A Jupyter notebook (.ipynb) under docs/demos/, wired into the docs navigation alongside the existing demos.

Scope

• [[7,1,3]] 2D color code (Steane code) curve
• [[15,1,3]] 3D color code (quantum Reed-Muller code) is only required if Tsim can run the circuit with reasonable compute, since this circuit requires 15 non-Clifford rotations. You may also want to experiment with clifft.
• Curve for unentangled experiment
• Explain the physics — include markdown that explains global rotations, the codes being studied, and what tsim is computing at each step, so the notebook reads as a tutorial and not just a script.

Considerations

The publication uses deterministic encoding circuits, described in the extended data (Fig 10a) of the paper. It can be useful to use the method stim.Circuit.flow_generators() to verify that your encoding circuit has the correct stabilizers and logical Pauli operators.

[ljcamargo]

Hi @rafaelha I just submitted a PR #155 for unitaryHack. As I said on the PR, could manage to get the 3d code working and still actively researching if this is because a computational constraint or other error, but do expect an update as soon as I find out. Thanks a lot!

[shalini9894]

Hi @rafaelha, I submitted a PR#156. I could able to create the graphs resembling as closely as possible to Fig 4a. I also ran Steane code and Reed Muller. Thanks so much. It was great working on this issue.

[shalini9894]

Hi @rafaelha, I submitted a PR#156. I could able to create the graphs resembling as closely as possible to Fig 4a. I also ran Steane code and Reed Muller. Thanks so much. It was great working on this issue.

I updated the graph. PR #156

[rosspeili]

Pushed an ideal (noiseless) Fig. 4a tutorial: docs/demos/global_rotations.ipynb.

Covers [[7,1,3]] (hypercube encoder + Stim flows), [[15,1,3]], unentangled 15-qubit baseline, and the negative-stabilizer control. Two-panel plot: logical X + X-stabilizer revival.

This is a noiseless tutorial, not the full experimental pipeline (decoder/postselection). The 15-qubit hypercube circuit isn’t in the SI, so 3D uses stabilizer prep instead.

Branch: https://github.com/rosspeili/tsim/tree/issue-119-global-rotations-tutorial