A Note On Lattice Definition of Gauge Theories With Chiral Fermions

summary

The gist

Lattice gauge theory with chiral fermions is described by interpolating a lattice gauge field over the continuum and coupling it to continuum chiral fermions to define a regularization scheme that…

In short

The paper develops a method to define chiral fermion path integrals on arbitrary manifolds by interpolating admissible lattice gauge fields with continuum chiral fermions. This approach relies on the well-definedness of continuum path integrals and uses concepts like eta-invariants and quantum renormalized theta-angles to define partition functions, enabling numerical implementation.

Key concepts

Admissible Lattice Gauge Fields
These are lattice gauge fields whose plaquette holonomies are close to the identity (i.e., |UP - 1| is small). They allow for a smooth interpolation from the discrete lattice structure to a continuum gauge field, which is crucial for coupling with continuum chiral fermions without introducing fermion doubling problems.
Fermion Path Integral Formula
The partition function is characterized by $| ext{det } D/ + W | ext{exp}(-i ilde{ heta})$, where $D$ is the Dirac operator, $W$ is the gauge field, and $ ilde{ heta}$ relates to the eta-invariant. This formula remains valid for any anomaly-free representation and depends only on the spacetime manifold W and the gauge field.
Quantum Renormalized Theta-Angles
These angles are associated with homomorphisms from four-dimensional cobordism groups to U(1). They are essential for defining the chiral fermion path integral on an arbitrary manifold W, allowing for a general definition that does not depend on the specific five-manifold Y used in the construction.
Eta-Invariant ($\eta(D_Y)$)
The eta-invariant is computed using the spectrum of a self-adjoint operator $D_{2Y}$ (a Laplace-like operator) with generalized APS boundary conditions. This method avoids direct lattice regularization issues and allows for the computation of $\eta(D_Y)$, which is then used in the partition function formula.

Terminology used across episodes

This episode discusses

The paper

A Note On Lattice Definition of Gauge Theories With Chiral Fermions · Read on arXiv

Edward Witten

School of Natural Sciences, Institute for Advanced Study

Transcript

Introduction to the show: ident: High Energy Physics: Lattice. The day in lattice field theory: what the simulations put on the table.

Kai: We're looking at the paper: "A Note On Lattice Definition of Gauge Theories With Chiral Fermions".

Mira: Here's the gist of the paper. Lattice gauge theory with chiral fermions is described by interpolating a lattice gauge field over the continuum and coupling it to continuum chiral fermions to define a regularization scheme that can be implemented in principle on a computer <ref:2611.05051#pg33> The paper details an approach based on interpolating an admissible lattice gauge field over the continuum and coupling it to continuum chiral fermions, which provides a method for defining the fermion path integral in four dimensions for any anomaly-free representation <ref:2611.05051#pg8> This technique relies on two key facts: the well-definedness of the fermion path integral in continuum Yang-Mills theory and its ability to be implemented numerically by coupling it to an interpolated lattice gauge field <ref:2611.05051#pg4> The analysis generalizes previous results, applying to any gauge group and anomaly-free fermion representation in any dimension <ref:2611.05051#pg4> The Well-Definedness of the Continuum Path Integral The first key fact is that in continuum Yang-Mills theory with any gauge group G, the path integral for fermions in any anomaly-free representation, chiral or not, coupled to any classical background field A (and any background spacetime metric), is always well-defined <ref:2611.05051#pg4> This statement is supported by two points: 1.

Kai: In learning how to define the fermion path integral for anomaly-free theories, one also learns to determine all possible nonperturbative anomalies <ref:2611.05051#pg5>.

Mira: For example, in four-dimensional chiral gauge theories with connected and simply-connected gauge group, one learns that the only anomaly beyond the usual triangle anomalies is the familiar mod 2 anomaly involving the number of fermion zero-modes in an instanton field <ref:2611.05051#pg5>.

Title and authors: Kai: We’ve been talking about the core idea behind "A Note On Lattice Definition of Gauge Theories With Chiral Fermions," specifically how they regularize chiral gauge theories on a lattice using interpolated fields and continuum fermions.

Mira: They are really tackling the fermion doubling problem head-on by using interpolation, which is crucial because it allows them to couple lattice gauge fields to continuum chiral fermions without getting that problem in the first place <ref:2611.05051#pg4>.

Lev: From an error correction standpoint, I see the value of that interpolation method being in how much overhead we’d need for those iterative Landau gauge problems <ref:2611.05051#pg3>.

Kai: The paper also points out that they can make improvements on this approach by using a specific formula derived from anomaly inflow and eta-invariants to define the lattice path integral, which is something they call equation one point two <ref:2611.05051#pg4>.

Mira: That formula is important because it allows them to incorporate the discrete and continuous theta-angles into the lattice path integrals by relating them to homomorphisms from the cobordism group to U(one) <ref:2611.05051#pg9>.

Lev: So, for us, it means that if we can successfully implement that on hardware, we’d be dealing with an integral over a compact space of finite dimension <ref:2611.05051#pg4>.

Kai: Right, and the paper also suggests improvements in how they compute the absolute value of the fermion determinant by using a Laplace-like operator D2Y with an inherited APS boundary condition to avoid doubling in numerical study <ref:2611.05051#pg4>.

Mira: That sounds like a neat trick for handling the absolute value of the determinant, which is usually hard because it avoids the positivity issue in Monte Carlo methods <ref:2611.05051#pg4>.

Lev: If we can compute that spectral information from D2Y without doubling, that drastically reduces what we have to worry about when translating this to a real system <ref:2611.05051#pg4>.

Kai: So the overall message is they are providing a path from the continuum definition of the fermion path integral to an implementable lattice scheme, even if it’s not suitable for standard Monte Carlo methods <ref:2611.05051#pg1>.

Mira: It shifts the focus from being stuck on the doubling problem to defining how to handle these nonperturbative aspects through this regularization scheme <ref:2611.05051#pg4>.

Lev: It changes the overhead landscape because it moves us away from needing a full positive measure, but we still have that finite dimension integral to manage <ref:2611.05051#pg4>.

The paper's summary: Kai: To summarize what they found, the paper is building on previous work by Craig and Hsu to describe a lattice regularization of chiral gauge theories that applies in any dimension and for any anomaly-free representation <ref:2611.05051#pg3>.

Mira: They use two key facts: the first is the well-definedness of the fermion path integral in continuum Yang-Mills theory and its ability to be implemented numerically by coupling it to an interpolated lattice gauge field <ref:2611.05051#pg4>.

Lev: So they are essentially saying that because we know how to do it in the continuum, we can use that knowledge as a basis for defining the path integral on the lattice <ref:2611.05051#pg4>.

Kai: The second fact is that this approach works because it uses no special properties of the gauge group or fermion representation, making it applicable to any anomaly-free representation in four dimensions <ref:2611.05051#pg3>.

Mira: And they also explain the path integral formula for an anomaly-free system, chiral or not, can be described by explicit formulas that could be evaluated numerically on a computer <ref:2611.05051#pg5>.

Lev: So the main result here is that they provide an explicit formula for the fermion partition function involving the absolute value of a determinant and an exponential term related to the eta-invariant <ref:2611.05051#pg4>.

Kai: That formula is valid in any anomaly-free representation and depends only on the spacetime W and the gauge field <ref:2611.05051#pg5>.

Mira: It connects everything together by showing how to define the path integral on an arbitrary manifold W by considering its cobordism structure and using quantum renormalized theta-angles <ref:2611.05051#pg9>.

Lev: So they are giving us a general definition that doesn't depend on picking a specific five-manifold Y, just the manifold W <ref:2611.05051#pg9>.

The paper's improvements: Kai: The paper suggests improvements center around using the interpolation of admissible lattice gauge fields to construct a smooth continuum gauge field by acting on that interpolated field with the Yang-Mills gradient flow for a short time <ref:2611.05051#pg4>.

Mira: That smoothing step is important because it ensures the continuum gauge field is of order epsilon in lattice units, which helps ensure we have a good continuum limit <ref:2611.05051#pg4>.

Lev: From an engineering view, that gradient flow step means we’re solving finitely many Landau gauge Dirichlet problems which can be solved numerically <ref:2611.05051#pg4>.

Kai: And then they also discuss how to define smooth continuum gauge fields by acting on the interpolated A with that gradient flow for a short time <ref:2611.05051#pg4>.

Mira: This is a concrete way to get a smooth continuum gauge field in the same topological class as the original lattice configuration <ref:2611.05051#pg4>.

Lev: It sounds like we’re proposing an iterative numerical procedure that builds the field piece by piece, which is something I can actually build for real hardware <ref:2611.05051#pg4>.

Kai: And then they also have this idea about constructing smooth continuum interpolations of admissible nonabelian lattice gauge fields using a sequence of increasingly fine lattices where the top dimensional cells have bounded flatness to ensure a continuum limit <ref:2611.05051#pg4>.

Mira: That bottom part is what ensures we can actually get that continuum limit you need for your experiments, which is essential for any experimental setup <ref:2611.05051#pg4>.

Conclusion: Kai: So to wrap up this paper on "A Note On Lattice Definition of Gauge Theories With Chiral Fermions," they are showing that the main result is that the fermion partition function on any manifold W with a boundary condition L along W equals partial Y is det D/ plus W exp of negative i pi eta <ref:2611.05051#pg4>.

Mira: And they emphasize that this formula is independent of your choice of the five-manifold Y and depends only on the manifold W and the fields on it <ref:2611.05051#pg9>.

Lev: This is a big deal because it means we’s not stuck needing to pick a specific five-manifold Y to define the integral <ref:2611.05051#pg9>.

Kai: It means you can use this general definition for any W in terms of the quantum renormalized theta-angles <ref:2611.05051#pg9>.

Mira: And they provide a general formula for the partition function involving those theta-angles, which shifts them, which is how you get the effect <ref:2611.05051#pg9>.

Lev: So it’s a way to define this path integral without running into those topological obstructions that obstruct your construction <ref:2611.05051#pg9>.

Kai: It’s a nonperturbative version of anomaly inflow where the inflow from the bulk compensates for all perturbative and nonperturbative anomalies along the boundary <ref:2611.05051#pg9>.

Mira: So for our next paper, we'll be looking at how they apply this framework to define observables on manifolds, which will give us a clearer picture of what these new tools can actually do <ref:2611.05051#pg9>.

Lev: I just want to say that the final result for the partition function, ZPsi,APS equals one when m equals minus M <ref:2611.05051#pg9>.

Kai: That's a key piece of information about how the phase behaves, showing Arg of ZPsi,APS can be written as X k Arg minus i lambda k plus m minus i lambda k plus M <ref:2611.05051#pg9>.

Mira: So for anyone listening, this paper is giving us a really solid mathematical framework to define the fermion path integral on complicated spaces without relying on a specific geometry <ref:2611.05051#pg9>.

More episodes

← Home