Quantum topology, a rapidly evolving field, explores the potential of quantum mechanics to create and manipulate topological states. This paper introduces a novel algorithm, Quantum Topology Transformation Learning (QTTL), designed to learn complex topological structures through the application of high-order time integrals. QTTL leverages this property to dynamically adapt to structural changes, achieving efficient and automated topological transformations. The core mechanism centers on exploiting the inherent temporal dynamics within high-order time integrals to guide the optimization process, ultimately offering a departure from traditional approaches that often rely on manual design. This work proposes a new framework for automatic topology exploration and transformation, potentially offering significant advancements in the study of topological materials and phenomena.
In this work, we present some new understanding of topological order, including three main aspects. (1) It was believed that classifying topological orders corresponds to classifying gapped quantum states. We show that such a statement is not precise. We introduce the concept of gapped quantum liquid as a special kind of gapped quantum states that can ``dissolve'' any product states on additional sites. Topologically ordered states actually correspond to gapped quantum liquids with stable ground-state degeneracy. Symmetry-breaking states for on-site symmetry are also gapped quantum liquids, but with unstable ground-state degeneracy. (2) We point out that the universality classes of generalized local unitary (gLU) transformations (without any symmetry) contain both topologically ordered states and symmetry-breaking states. This allows us to use a gLU invariant---topological entanglement entropy---to probe the symmetry-breaking properties hidden in the exact ground state of a finite system, which does not break any symmetry. This method can probe symmetry- breaking orders even without knowing the symmetry and the associated order parameters. (3) The universality classes of topological orders and symmetry-breaking orders can be distinguished by stochastic local (SL) transformations (i.e., local invertible transformations): small SL transformations can convert the symmetry-breaking classes to the trivial class of product states with finite probability of success, while the topological-order classes are stable against any small SL transformations, demonstrating a phenomenon of emergence of unitarity. This allows us to give a definition of long-range entanglement based on SL transformations, under which only topologically ordered states are long-range entangled.
In fault-tolerant quantum computing with the surface code, non-Clifford gates are crucial for universal computation. However, implementing these gates is usually more challenging and resource-intensive than Clifford gates. Methods such as distilling a high-fidelity magic state from noisy copies or transforming to higher-dimensional codes require a significant qubit count overhead. In this work, we propose a new protocol that combines magic state preparation and code transformation to realize logical non-Clifford operations, which has the potential for reducing the qubit count overhead, provided an efficient decoding algorithm exists for the intermediate non-Abelian code involved. Our approach begins with a special logical state in the
Z4 surface code. By applying a sequence of transformations, the system goes through different topological codes, including the non-Abelian D 4 quantum double model. This process ultimately produces a magic state encoded in the
Z2 surface code. A logical T gate can be implemented in the standard
Z2 surface code by gate teleportation. In our analysis, we employ a framework where the topological codes are represented by their topological orders and all the transformations are considered as topological manipulations such as gauging symmetries and condensing anyons. This perspective is particularly useful for understanding transformations between topological codes.