K K -theoretic computation of the Atiyah(-Patodi)-Singer index of lattice Dirac operators

Shoto Aoki, Hidenori Fukaya, Mikio Furuta, Shinichiroh Matsuo, Tetsuya Onogi, Satoshi Yamaguchi · 2025

Read K K -theoretic computation of the Atiyah(-Patodi)-Singer index of lattice Dirac operators by Shoto Aoki, Hidenori Fukaya, Mikio Furuta, Shinichiroh Matsuo, Tetsuya Onogi, Satoshi Yamaguchi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-23921.txt.

Introduction

Shoto Aoki, Hidenori Fukaya, and four colleagues compute a famous topological index for lattice Dirac operators using K-theory, in a paper published in two thousand twenty-five. The index counts a mismatch between left- and right-handed zero modes, and the Atiyah-Singer theorem ties it to the geometry of the underlying space.

Their approach treats the Wilson Dirac operator of lattice gauge theory as an object in K-theory, and shows its spectral flow equals the index. Unlike the standard lattice formulation, theirs needs no Ginsparg-Wilson relation. It also extends to systems with boundaries and to mod-two versions of the index in general dimensions. Numerical checks confirm the formulas reproduce the known continuum theorems. For the first time, they compute the Atiyah-Patodi-Singer index with curved boundaries, and its mod-two cousins, directly on a lattice. The work gives lattice practitioners a cleaner topological toolkit.

Catalog record

CREATORShoto Aoki, Hidenori Fukaya, Mikio Furuta, Shinichiroh Matsuo, Tetsuya Onogi, Satoshi Yamaguchi
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-23921

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

K K -theoretic computation of the Atiyah(-Patodi)-Singer index of lattice Dirac operators Shoto Aoki, Hidenori Fukaya, Mikio Furuta, Shinichiroh Matsuo, Tetsuya Onogi, Satoshi Yamaguchi Abstract We show that the Wilson Dirac operator in lattice gauge theory can be identified as a mathematical object in K K -theory and that its associated spectral flow is equal to the index. In comparison to the standard lattice Dirac operator index, our formulation does not require the Ginsparg-Wilson relation and has broader applicability to systems with boundaries and to the mod-two version of the indices in general dimensions. We numerically verify that the K K and K ​ O KO group formulas reproduce the known index theorems in continuum theory. We examine the Atiyah-Singer index on a flat two-dimensional torus and, for the first time, demonstrate that the Atiyah-Patodi-Singer index with nontrivial curved boundaries, as well as the mod-two versions, can be computed on a lattice. † † preprintnumber: OU-HET-1267 1 Introduction The index of Dirac operators [ 1 ] has played an important role in particle theory to understand nonperturbative nature of gauge theories. It is closely related to chiral symmetry as well as to gauge field topology, which is a property under continuous deformation of fields on a continuous spacetime. However, its formulation in lattice gauge theory, where chiral symmetry is broken and spacetime is discretized, has been a challenging problem. It is well-known that the overlap Dirac operator [ 2 ] and the operator in the perfect action [ 3 ] achieve an exact chiral symmetry through the Ginsparg-Wilson relation [ 4 ] and the associated Atiyah-Singer(AS) index is well-defined. The formulation is, however, limited to a periodic/antiperiodic square lattice whose continuum limit is a flat torus. Recently the authors gave a mathematical proof that a family of the massive Wilson Dirac operators on an even-dimensional square lattice can be identified as a K group element [ 5 , 6 ] defined on the mass parameter space and its continuum limit given by the \eta invariant or equivalently the spectral flow [ 7 ] 1 1 1 The equivalence of the spectral flow of the Wilson Dirac operator to the index was empirically known in [ 8 ] and shown in [ 9 ] to the overlap Dirac index. Its mathematical meaning or relation to the K -theory was, however, not discussed. , converges to the index of the continuum Dirac operator. In contrast to the standard overlap index, our new K -theoretic treatment of the lattice Dirac operator does not rely on the Ginsparg-Wilson relation at all, and offers wider applications to the systems where chiral symmetry is absent or difficult to realize. One such system is a manifold with boundaries in which the Ginsparg-Wilson relation is generally broken [ 10 ] . The K -group element defined by the Wilson fermion Dirac operator can be straightforwardly extended to the case with open boundary conditions, which is equivalent to the domain-wall fermions [ 11 , 12 ] . In [ 13 , 14 ] , in which five of the authors were involved, it was proved in continuum theory that the Atiyah-Patodi-Singer (APS) index [ 15 ] of the Dirac operator is equal to the \eta invariant of the massive Dirac operator where the mass term has domain-walls at the location of the original boundaries. A perturbative equivalence between the continuum and lattice \eta invariants was examined in Ref. [ 16 ] and a more rigorous mathematical justification will be given in [ 17 ] . Another example is the mod-two index [ 18 ] of real Dirac operators, which appears in arbitrary dimensions. A well-known example is the one in the SU(2) gauge theory in five dimensions, which is known as the origin of the Witten anomaly [ 19 ] . In odd dimensions, chiral symmetry is absent, and the overlap Dirac operator is not directly available. In contrast, K -theory still works with modifications that are well-known in mathematics: switching from K groups to KO groups. These results are nontrivial in mathematics (for different mathematical approaches, we refer to Refs. [ 20 , 21 ] ). The index of the Dirac operator is generally formulated and nontrivial only when the system has an infinite-dimensional Hilbert space. Our formulation may lead to a systematic way of a finite-dimensional approximation of topological invariants of general vector bundles. We also remark that our K -theoretic formulas for the APS index provide a natural connection between the massive bulk and the massless edge-localized fermions, which may be useful for understanding topological insulators and superconductors. In this work, we numerically verify that our K -theoretic formulas yield values consistent with the index theorems in continuum theory. On a two-dimensional square lattice, we first examine the Atiyah-Singer index theorem for a U(1) gauge theory on a flat torus T^{2} . Then we introduce nontrivial domain-walls, including curved ones [ 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 ] , and investigate the Atiyah-Patodi-Singer index theorem, in which the edge-localized modes at the domain-wall play a crucial role. We also study a Majorana fermion system with a boundary to test the mod-two index formula. To the best of our knowledge, this is the first numerical study demonstrating that the APS index and its mod-two version can be computed on a lattice. 2 K -theoretic formulas Here we summarize the index formulas obtained from K -theory. For mathematical details, see [ 7 , 17 ] . We also provide a brief summary in Appendices. A and B . In the standard formulation of the index in continuum theory, the massless Dirac operators are interpreted to represent the K^{0}(\{0\}) or KO^{-1}(\{0\}) group elements. The basic idea is to add and vary the mass parameter by -sM where M is positive, and s\in I=[-1,1] . We denote its two end points by \partial I=\{-1,1\} . The original zero eigenvalues of the massless Dirac operator cross zero along the path I . Thus the index can be evaluated appropriately counting the crossing-zero modes, or the spectral flow. The equivalence is mathematically guaranteed by the so-called suspension isomorphism between the K^{0}(\{0\}) and K^{1}(I,\partial I) groups (or KO^{-1}(\{0\}) and KO^{0}(I,\partial I) groups for real Dirac operators.). For a general complex Dirac operator we take \gamma(D-sM) with s\in I , as a K^{1}(I,\partial I) group element under a condition that \gamma(D\pm M) have no zero mode. Here the chirality operator \gamma and its anticommutation relation with D are not essential and we do not have to distinguish the continuum and lattice Dirac operators in the formulation. Therefore, the massive Wilson Dirac operator (family) \gamma(D_{W}-sM) with 0<M<2 (to avoid fermion doubling) on an even-dimensional periodic square lattice can also be identified as an element of K^{1}(I,\partial I) . Denoting the number of zero-crossing modes from positive to negative by n_{+} and that from negative to positive by n_{-} , the spectral flow n_{+}-n_{-}=-\frac{1}{2}\eta(\gamma(D_{W}-M))+\frac{1}{2}\eta(\gamma(D_{W}+M)), (1) is well-defined and agrees with the overlap Dirac index. Here the eta invariant \eta(H) is difference between the positive and negative eigenvalues of a Hermitian operator H . Its equivalence to that in continuum theory at a sufficiently small lattice spacing was

Read the complete text