Daily Summary for 2026-11-15

daily

Transcript

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

Kai: It's the fifteenth of November, twenty twenty-six, and this is the day's research.

Mira: 9 new papers came out today.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We'll take the day in one pass, then pull out the papers we're staying with.

Mira: Start us off with Constrained field-space estimators for background-field lattice QCD observables.

The day's research: Kai: Today we focus on selecting models and data subsets when fitting correlated data without regularization methods.

Mira: This is important because standard methods often depend on the regularization used for the covariance matrix.

Lev: The RIAIC criterion estimates the model-dependent part of the correlated Kullback--Leibler divergence.

Kai: This lets researchers select subsets even when regularizations are applied to the fitted dataset's covariance matrix.

Mira: This idea draws on lattice quantum chromodynamics where treating symmetries carefully is essential.

Lev: In that field, researchers renormalize correlators using a common reference point in the ud sector to study emergent chiral-spin symmetry.

Kai: That ensures the resulting symmetry ratio stays physically meaningful across different channels.

Mira: Another key area involves extracting nucleon properties from lattice QCD where choosing how to examine correlation functions is vital.

Lev: This work builds optimized operators using distillation profiles to better overlap with the nucleon ground state, reducing signal-to-noise degradation.

Kai: Precision measurements of nucleon charges are improving through detailed lattice calculations using twisted-mass clover-improved fermions with physical quark masses.

Mira: These studies show good agreement for the isovector charge compared to other lattice calculations.

Lev: A regularization-independent Information Criterion for correlated data generalizes AIC to allow model and data subset selection when regularized fits are performed on correlated data.

Kai: Renormalized symmetry probes for emergent symmetries and non-chiral lattice fermions use renormalization to make symmetry ratio measurements independent of the specific regularization scheme.

Mira: Optimized Distillation for Nucleon Correlation Functions builds better operators by modulating Laplacian eigenmodes to improve the overlap with the nucleon ground state.

Lev: Symanzik Effective Hamiltonian for Lattice Quantum Field Theories proposes a framework based on dimensional analysis and symmetries to classify lattice operators at a given order in lattice spacing.

Kai: Unpolarized proton PDF with physical quark masses and large momenta presents a lattice QCD calculation of the proton's parton distribution function using large-momentum effective theory.

Mira: The structure of the pion at large N c computes the pion's momentum fraction in the large N c limit to show that gluons carry most of its momentum.

Lev: Reconstruction of pion polarizabilities in lattice QCD with machine learning uses a denoising diffusion model to reconstruct electric and magnetic polarizabilities from lattice data.

Kai: Flavor decomposition of the nucleon axial and tensor charges from lattice QCD determines nucleon charges using four ensembles of twisted-mass fermions to benchmark BSM interactions.

Mira: The flavor decomposition of the nucleon axial and tensor charges from lattice QCD determines nucleon charges using four ensembles of twisted-mass fermions to benchmark BSM interactions.

Lev: That is the last piece we reviewed today.

Lucky paper: 2611.10773: Kai: So we're looking at this paper, Constrained field-space estimators for background-field lattice QCD observables. It seems to be about how to extract weak-field response coefficients from those correlators computed at different field strengths.

Mira: It sets up this whole extraction problem from a numerical analysis perspective where the desired response coefficient is treated like a derivative or a coefficient of a specified weak-field basis evaluated at zero field.

Lev: That setup makes the field redundancy explicit, allowing additional field values to impose further response mode constraints, which is where you get into the constrained estimator formulation.

Kai: If you have that full set of N fields, and you only care about K modes, the linear estimator has weights where X i w i equals one to preserve the target coefficient.

Mira: But then to annihilate each selected nuisance mode, you need X i w i phi k(B) to be zero for all those modes from one to K minus one.

Lev: The minimum variance estimator within that constraint family is found by minimizing the weight vector w subject to the constraint A T w equaling b.

Kai: That leads to a best linear unbiased estimator, or BLUE, with weights w equal to sigma-one A A T sigma-one b, provided the matrix sigma is nonsingular and your constraints are full rank.

Mira: The paper then discusses the role of covariance optimization and omitted-mode leverage, which shows that if the true reduced response has an additional mode, say alpha phi star, its contribution to the estimator is measured by w T phi star.

Lev: If that leverage is large, a statistically stable minimum variance solution can still be systematically fragile because of Peelle’s pertinent puzzle.

Kai: That’s a big warning for anyone trying to deploy these extraction methods in production settings or for high-precision physics where systematic errors matter.

Mira: Moving on to the fixed field-space hierarchy, they compare fixed estimators with conventional response fits, showing that the fixed O4 stencil is algebraically related to conventional fits.

Lev: This formulation makes the n minus K remaining estimator degrees of freedom explicit, which is a key distinction from just using a standard fit.

Kai: The null space statistic tests bound discrepancies among estimators in the same constraint family, and for the O4 case with four fields, it has two degrees of freedom.

Mira: They found that for the O4 stencil, when you include temporal noise, it gets rejected for every ensemble except the lightest neutron at p equals zero point zero five three.

Lev: The covariance-optimized estimators then have a displacement from the fixed O4 stencil that is about one marginal standard error below the fixed O4 result for the proton moments at five lighter masses.

Kai: So, even with optimization, you are looking at one to two percent precision for those lighter masses and seven to eight percent precision at the heaviest mass when using covariance optimization.

Mira: And they mention that decuplet correlators provide targets known exactly from lattice contractions where the higher-order response can displace the lowest-order estimate by up to twenty-five standard errors.

Lev: That's a very concrete target for testing these methods, because you have something known exactly to compare against the complex extraction.

Kai: Data and code availability is covered too, they archive all derived field-response vectors and scripts at Zenodo so people can run the full pipeline.

Mira: The underlying correlators themselves are also available from the authors if you make a reasonable request, which is pretty open for verification.

Lev: This work was supported in part by the U.S. Department of Energy under Grant No DEFG02-95ER40907 and most of the heavy calculations were done at NERSC.

Lucky paper: 2611.10816: Kai: So we're diving into the paper on Flavor decomposition of the nucleon axial and tensor charges from lattice QCD, which uses four ensembles with twisted-mass clover-improved fermions.

Mira: That paper is tackling how to determine the axial and tensor charges for each quark flavor f, denoted as g f A,T.

Lev: The authors derive these by looking at nucleon matrix elements of corresponding axial and tensor operators at zero momentum transfer, where they are defined by that specific expression involving u-bar N gamma A,T u N.

Kai: They use twisted-mass fermion discretization which gives automatic O(a) improvement and adds a clover term to suppress isospin-breaking effects.

Mira: To get those matrix elements, they need both two-point and three-point Euclidean time correlation functions.

Lev: The two-point function is written as C of Gamma zero vector p times t s times x sub s minus i vector x sub s dot p times Tr h Gamma zero A N t s minus J N zero at the origin.

Kai: They use Gaussian smearing on the quark fields going into the interpolating field to improve the overlap with the true nucleon ground state and suppress excited-state contamination.

Mira: That leads to a three-point function, C mu of Gamma k vector q times p prime t s times t ins times x sub s minus i vector x sub s dot q at t ins, which they then analyze using spectral decomposition.

Lev: To isolate the matrix elements from those unknown overlaps, they build a ratio R mu A of the three-point function to the two-point function.

Kai: In the limit of large time separations, where Delta E between the first excited state and ground state is much larger than one, this ratio R mu A gives them that desired axial charge gA.

Mira: They also get a ratio R mu nu T for the tensor operator which yields the tensor charge gT, defined using an epsilon i j k symbol times T zero zero ij Gamma k c sub zero.

Lev: For renormalization, they use nonperturbative renormalization in the RI prime minus MOM scheme followed by perturbative conversion to the MS scheme at a reference scale of two GeV.

Kai: They get their continuum limit by using a model average of constant and linear in a squared lattice spacing fits for that extrapolation.

Mira: Importantly, they perform the continuum extrapolation using ensembles simulated at approximately the physical pion mass, which avoids needing a chiral extrapolation.

Lev: The paper compares their results with other recent lattice QCD determinations, finding good agreement for the isovector axial charge and flavor-diagonal charges.

Kai: They note that for the up-quark axial charge specifically, their result is smaller than what a previous calculation from PNDME gave.

Mira: The summary shows they determined the nucleon charges using four Nf equals two plus one plus one twisted-mass clover-improved ensembles with quark masses tuned close to physical values.

Lev: The final continuum results show g u A equals zero point eight six zero minus zero point four oh four minus zero point zero three six, and for the tensor charge, g u T is zero point nine five one minus...

Kai: They also found good consistency for the flavor-diagonal axial and tensor charges across different lattice calculations.

Mira: The systematic uncertainty on the continuum value comes from investigating excited states where they use a summation method for connected contributions and a weighted average for disconnected ones.

Lev: For example, their final result for the up quark axial charge is zero point eight nine five minus zero point four two zero minus zero point zero three four, which carries a statistical uncertainty of seventeen percent.

Kai: The continuum extrapolation has a non-negligible effect on the physical values even though the lattice spacing dependence is mild.

Mira: They are careful to avoid chiral extrapolations by simulating at the physical pion mass, which keeps things cleaner.

Lev: It sounds like they have a very solid set of results, especially since their isovector axial charge matches both other lattice results and experimental data.

Lucky paper: 2611.12520: Kai: Alright, we're moving into our third paper of three today, which is the Symanzik Effective Hamiltonian for Lattice Quantum Field Theories.

Mira: This paper proposes a framework to construct that effective Hamiltonian directly using dimensional analysis and symmetries in the Hamiltonian formulation.

Lev: It allows researchers to classify all operators that can contribute at a given order in the spatial lattice spacing by using those same dimensional analysis and symmetries of the theory, which you can reduce further with equations of motion and other identities.

Kai: So this framework gives a Hamiltonian counterpart to Symanzik’s effective action, but it avoids dealing with higher order time derivatives because you handle those first.

Mira: The paper shows an explicit correspondence for scalar field theory where the construction for the tree-level expansion of the Hamiltonian is checked at order two spatial lattice spacings.

Lev: At order four spatial lattice spacings, they give a list of allowed operators in Hermitian ordering including terms like grad two phi and phi three pi.

Kai: It mentions that there's freedom to add time derivatives to the second-order Hamiltonian, and you can remove those redundancies by using commutators with the leading continuum Hamiltonian.

Mira: For pure lattice gauge theory, they define the KS lattice Hamiltonian whose structure is constrained by Gauss' law, and relevant symmetries include parity P, charge conjugation C, and time reversal T.

Lev: The list of allowed Hermitian operators for the dimension-six contribution to H2 in pure SU(N) gauge theory includes M1, M3, E1, K1, and K2.

Kai: This matching reproduces known quadratic corrections to the Kogut–Susskind Hamiltonian when you check it at tree level.

Mira: The paper discusses the connection between this effective Hamiltonian and Symanzik’s effective action through a Legendre transformation applied to an anisotropic lattice via a temporal continuum limit.

Lev: They use this route as a complementary way to get the description of the Hamiltonian once those higher order time derivatives have been handled.

Kai: So it's building a basis for an improvement programme starting from whatever lattice Hamiltonian you are given right now.

Mira: The paper concludes by applying this framework to both scalar field theory and pure gauge theory, constructing operator bases up to order two spatial lattice spacings.

Lev: They state that the classification excludes corrections linear in the spatial lattice spacing for all orders in perturbation theory because of how they define the terms.

Kai: What are some of the specific examples of those operators they found for scalar field theory at order four?

Mira: They listed grad two phi and X i(d two i phi) squared, along with grad pi squared and phi three pi.

Lev: For pure gauge theory, they look at the dimension-six operators built from E1, B1, and terms like F ij = epsilon ijkB k.

Kai: I see how this helps simplify things by focusing on what matters at a given lattice spacing.

Mira: The paper also mentions that the direct Hamiltonian construction avoids dealing with higher order time derivatives while the effective action route requires them to be treated beforehand.

Lev: This framework aims to provide a basis for improvement programmes, but they note that developing non-perturbative improvement conditions and completing the non-Abelian interaction matching are subjects for future research.

More episodes

← Home