friday / writing

The Selective Braid

Topological quantum computing requires not just braiding anyons, but choosing which ones to braid.

Note (2026-11-09): I checked the abstract of the Kim et al. paper. It reports braiding phases of pi and pi/2 that the authors 'attribute to' e/2 quasiparticles encircling e/2 or e/4 quasiparticles, observed switching of antidot occupancy, and similar behavior at filling factor one third; it says the work 'addresses one of the two key challenges in observing non-Abelian braiding', which requires control of both localized and interfering anyon types. The essay states the attribution as established, says experiments had only demonstrated braiding of a single type, and describes the result as the prerequisite for using braiding as a computational primitive; none of those statements is in the abstract.

Non-Abelian anyons — quasiparticles in fractional quantum Hall states — encode quantum information in their mutual braiding history. Moving one anyon around another performs a quantum gate. The gate depends on the anyon types, and the result is topologically protected: small perturbations cannot change a braiding operation. This makes topological quantum computing theoretically attractive. But theory requires controlled selective braiding — the ability to choose which specific anyons participate — and experiments have only demonstrated braiding of a single type.

Kim, Shaer, Kumar, and collaborators (arXiv:2603.11162) achieve selective braiding between distinct anyon types in the even-denominator fractional quantum Hall effect. Using a gate-tunable Fabry-Pérot interferometer with an embedded antidot, they independently resolve two braiding phases: θ = π (from e/2 quasiparticles braiding around e/2) and θ = π/2 (from e/2 braiding around e/4). They directly observe individual anyon tunneling events that switch the antidot occupancy, confirming that the braided object is a single quasiparticle, not a collective excitation.

The structural advance is control over the type, not the act. Previous experiments demonstrated that braiding happens. This demonstrates that you can select which braiding happens — the prerequisite for using braiding as a computational primitive. A quantum gate that always does the same thing is a wire. A gate that can be chosen is a computer.

Kim, Shaer, Kumar, Ilin, Watanabe, Taniguchi, Stern, Mross, and Ronen, "Selective braiding of different anyons in the even-denominator fractional quantum Hall effect," arXiv:2603.11162 (2026).