Topological Preparation of Non-Stabilizer States and Clifford Evolution in $SU(2)_1$ Chern-Simons Theory

We develop a topological framework for preparing families of non-stabilizer states, and computing their entanglement entropies, in $SU(2)_1$ Chern-Simons theory. Using the Kac-Moody algebra, we construct Pauli and Clifford operators as path integrals over 3-manifolds with Wilson loop insertions, enabling an explicit topological realization of $W_n$ and Dicke states, as well as their entanglement properties. We further establish a correspondence between Clifford group action and modular transformations generated by Dehn twists on genus-$g$ surfaces, linking the mapping class group to quantum operations. Our results extend existing topological constructions for stabilizer states to include families of non-stabilizer states, improving the geometric interpretation of entanglement and quantum resources in topological quantum field theory.

💡 Research Summary

**
This paper develops a comprehensive topological framework for preparing and analyzing families of non‑stabilizer quantum states within the three‑dimensional SU(2)₁ Chern‑Simons theory. The authors begin by reviewing how entanglement entropy can be expressed in a Chern‑Simons setting using a single‑replica construction: the entropy of a subsystem A is obtained from a ratio of partition functions on a doubled manifold with an exchange diffeomorphism acting on the boundary tori associated with A.

The core technical contribution is the explicit construction of the Pauli and Clifford groups as topological operators. By exploiting the level‑1 Kac‑Moody algebra of SU(2), the authors identify the fusion matrix N as the Pauli‑X operator (shifting the label a → a + 1 mod 2) and the modular S‑matrix as the Hadamard (quantum Fourier transform) that conjugates X into Z. The modular T‑matrix supplies the phase gate. These three ingredients—S, T, and a controlled‑sum gate Cₛᵤₘ—generate the full Clifford group for qubits. Crucially, each gate is realized as a path integral over a three‑manifold with appropriately placed Wilson loops, making the quantum operations purely geometric objects.

With these tools, the authors present a topological protocol for preparing Wₙ states (equal superpositions of n‑qubit basis states with a single excitation) and, more generally, Dicke states. The preparation uses a Heegaard splitting of a genus‑n handlebody: each handle carries a Wilson loop in the fundamental representation, and the gluing of two copies of the handlebody implements the state preparation. By applying the replica trick—gluing two copies of the prepared manifold and inserting a Dehn twist on the boundary tori belonging to subsystem A—the reduced density matrix for A is obtained, and the von Neumann entropy is computed analytically. The result reproduces the expected linear scaling S(A) = |A| log 2 for stabilizer‑like partitions, but for Wₙ states the entropy scales as (n − |A|) log 2, reflecting the non‑stabilizer nature of the state.

The paper then establishes a precise correspondence between Clifford group actions and mapping‑class‑group (MCG) transformations. The S‑gate corresponds to the modular S transformation, the T‑gate to a Dehn twist, and the controlled‑sum gate to a composition of Dehn twists and handle slides that exchange two boundary tori. These topological moves act as normalizers of the Pauli operators, confirming that the Clifford group is isomorphic to the subgroup of the MCG generated by S and T. Consequently, any Clifford orbit of a non‑stabilizer state can be interpreted as a family of three‑manifolds related by MCG moves, and the entanglement entropy remains invariant across the orbit, demonstrating Clifford‑invariance of the entropy.

Finally, the authors discuss implications and future directions. By showing that non‑stabilizer resources such as Wₙ and Dicke states can be encoded purely in the topology of SU(2)₁ Chern‑Simons theory, they broaden the scope of topological quantum field theory as a platform for quantum information beyond stabilizer codes. The work suggests that even low‑level modular tensor categories possess sufficient structure to host complex multipartite entanglement, opening avenues for topological quantum computation that leverages non‑Clifford resources. Future research may explore higher‑level SU(2)ₖ theories or other gauge groups to realize richer families of non‑stabilizer states, and investigate experimental realizations of the required three‑dimensional manifolds and Wilson‑loop insertions in condensed‑matter or photonic platforms.