An Information-Theoretic Perspective on Spectral Reconstruction in Lattice QCD
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: High Energy Physics: Lattice. The day in lattice field theory: what the simulations put on the table.
Kai: I'm Kai.
Mira: And I'm Mira, here with Lev.
Kai: We're looking at the paper: "An Information-Theoretic Perspective on Spectral Reconstruction in Lattice QCD".
Mira: Here's the gist of the paper. Reconstructing spectral functions from finite, noisy lattice correlators requires understanding the constraints supplied by the data relative to prior assumptions This information-theoretic framework uses Kullback–Leibler (KL) divergences to quantify the information supplied by the data through changes from prior to posterior probability distributions The analysis accounts for both changes in the probability distributions of spectral strength at individual energies and constraints encoded in correlations across the spectrum Information-Theoretic Framework The framework compares the posterior density p with the prior density q using the KL divergence DKL(p∥q) = Z dx p(x) log2 p(x) q(x) This divergence measures the information supplied by the data relative to the prior The joint posterior of ρ and α is defined as p(ρ, α D) = p(D ρ) p(ρ α) p(α) The KL chain rule separates the information supplied by the data into an update of α and a spectral update at fixed α The energy-dependent measure K, with values Ki at ωi, measures the information supplied by the data about the spectral value at each energy relative to its conditional prior To include correlations, a full-conditional KL diagnostic J is introduced where Ji ≡ J(ωi) = Eα,ρ−iD h DKL p(ρi D, α, ρ−i)∥p(ρi α) This measure quantifies the average reduction in posterior uncertainty about ρi when the spectral values at all other energies are known, at fixed α Bayesian Reconstruction and Numerical Evaluation The reconstruction setup involves approximating the integral over a finite energy interval using Kti = ∆ωie−ωit, with ∆ωi the integration weight associated with ωi The spectral prior is defined by the BR functional T(ρ) = X i ∆ωi ρi mi− 1 − ln ρi mi, where mi = m(ωi) > 0 is the default model Posterior sampling in PyMC provides samples of both ρ and α while preserving their correlations The marginal posterior required for K is obtained by integrating over the remaining spectral values p(ρi D, α) = Z dρ−i p(ρi D, α, ρ−i) × p(ρ−i D, α) The information measures K and J are computed by averaging the KL divergence of each normalized conditional density in Eq.
Kai: (15) relative to its corresponding prior p(ρi α) over the joint posterior samples Mock-Data Results The mock spectrum contains three equal-weight peaks, ρtrue(ω) = X 3 n=1 δ(ω − En), (E1, E2, E3) = (1, 1.5, 3) The reconstruction captures the main spectral structures with centroids approximately 1.00, 1.49 and 2.94 The MAP selects the most probable spectrum, which reproduces the input peak positions and areas more faithfully than the posterior mean Information measures show that at the first peak, Ki ≃ 109.88 bits and Ji ≃ 117.43 bits Lattice-Correlator Results The analysis was applied to a proton two-point correlator calculated on the RBC/UKQCD 48I ensemble The information measures for the proton reconstruction show that K exhibits pronounced local dips near the two low-energy peaks The smoothed information curves provide a semiquantitative basis for assessing reconstruction reliability across energy regions In the high-energy region ω > 6 GeV, both smoothed measures decrease overall, indicating progressively weaker data constraints at higher energies Information-theoretic approaches offer a promising direction for analyzing lattice correlators and reconstructing spectral functions The conditional KL K quantifies the information supplied by the data through the change from the conditional prior to the posterior distribution of each spectral value The full-conditional KL J measures this information while also accounting for correlations with the rest of the spectrum Building on this study, future work will focus on using these developments to guide prior selection and quantify systematic uncertainties associated with prior choices Page 4 --- A variety of methods have been applied or developed to address this inverse problem, including the maximum entropy method (MEM)
14: , Bayesian reconstruction (BR)
15: , and the Backus-Gilbert (BG) method
13, 16: These approaches have been complemented by the Hansen-Lupo-Tantalo (HLT) method
17: and Chebyshev expansions
18: , as well as Gaussianprocess regression
19, 20: , sparse modeling
21: , and neural networks
22–24: Further developments include analytic formulations using conformal maps
25: and Mellin transforms
26: , together with convex optimization for bounding spectral integrals
27: Posterior sampling has also enabled Bayesian analyses beyond a single optimized spectrum
28: Page 5 --- The posterior distribution allows us to characterize the reconstructed spectrum through its mean, credible intervals and correlations Alongside the posterior mean, we consider the whole-spectrum maximum a posteriori (MAP) estimate, ρMAP = arg max ρ>0 p(ρ D), (14) We compare the posterior mean and MAP spectra in the numerical results Page 8 --- The local dips in K reflect its definition as a difference between posterior and prior distributions On either side of a peak, the reconstructed spectral strength changes between the small values in a valley and the large values near the peak center This suggests that the posterior distribution may also locally approach its conditional prior, providing a possible explanation for the dip in K For example, as shown in Fig.
Mira: 3, between the valley to the left of the first peak and its center, a dip occurs at ω ≃ 0.95, where Ki ≃ 0.05 bits Its proximity to the crossing of the posterior and prior means in Fig.
Kai: 1 is consistent with this interpretation Page 9 --- The MAP selects the most probable spectrum, whereas the posterior mean averages over the full posterior and is sensitive to its asymmetry and extended tails The choice of a spectral estimate depends on the properties one aims to recover
31: In this mock test, the MAP provides higher spectral resolution and reproduces the input peak positions and areas more faithfully than the posterior mean Although the mean reflects additional aspects of the posterior distribution, the MAP is the better choice for representing narrow spectral structures studied here Page 9 --- The upper panel of Fig.
Paper summary: Kai: So, we're looking at this paper called "An Information-Theoretic Perspective on Spectral Reconstruction in Lattice QCD." Basically, the authors are using information theory—specifically Kullback-Leibler divergence—to figure out how much data actually helps us reconstruct the spectral functions from noisy lattice correlators. It’s about quantifying exactly what kind of constraints that data is supplying compared to what we already assume.
Mira: Exactly. The main idea is framing this as a comparison between a prior distribution, which is our initial guess, and the posterior distribution, which includes the data information. They use this KL divergence to measure the information supplied by the data relative to that prior for both individual energy values and correlations across the whole spectrum.
Lev: From an engineering standpoint, that sounds like they're trying to map out where our input correlator data actually provides useful constraints versus where we just have a guess. How does this relate to what we're actually building?
Kai: Well, the paper sets up this framework by defining a joint posterior for the spectral density and some other parameter, and then they use that chain rule to separate the information supplied by the data into two parts: an update for that other parameter and an update for the spectrum itself at fixed values of that parameter.
Mira: And they introduce this measure K, which is essentially how much information we get about a single spectral value at one energy relative to its conditional prior, then they build on that with this joint diagnostic J, which accounts for correlations across all the other energies.
Lev: So it’s not just looking at one peak in isolation; they are trying to see how knowing what happens at all the other energies changes what we think is happening here. That's a significant step because real hardware always has these kinds of coupled constraints, right?
Kai: Right, and the paper shows how this works numerically using mock data—specifically three equal-weight peaks with positions one, one point five, and three. They show that their reconstruction method can capture those main spectral structures with centroids around one point zero zero, one point four nine, and two point nine four.
Mira: And they found that their maximum a posteriori estimate selects the most probable spectrum; it reproduces the input peak positions and areas more faithfully than just taking the posterior mean. That’s important because we don't always want to average everything out when reconstructing these things.
Paper summary: Lev: I see why that’s important for hardware deployment; if you need to identify a specific feature, you want the estimate that favors that feature, not some smoothed-out average. But what about how reliable those information measures are across different energy scales?
Kai: The results on the lattice correlators themselves show that the measure K exhibits local dips near those two low-energy peaks. These dips suggest a specific behavior where the posterior distribution might actually approach its conditional prior in certain regions, which is consistent with the way things look when you plot the posterior versus prior means.
Mira: That dip in K suggests that between a valley and a peak center, the reconstructed spectral strength changes between small values and large values, but it also gets closer to its conditional prior distribution. It’s a subtle point about how much the data is actually forcing us to change our initial assumptions.
Lev: So what does this imply for someone who just listens to the show? It means that even when the data seems strong, there are specific energy regions where it doesn't push us much beyond our starting assumptions. That points toward where we might need more precise input priors or more sophisticated data acquisition.
Kai: The paper also shows how these information measures behave at high energy, above six gigaevels. Both the smoothed measures K and J decrease overall as you go up in energy, falling to very small values, like on the order of ten to the power of negative three bits for K.
Mira: That decrease confirms what we expected from the mock study; it means that at higher energies, our data constraints are progressively weaker. The precision drops significantly because there's less information encoded in those high-energy correlators.
Lev: So when you think about running this on real hardware, you have to be very careful about how much weight you put on the high-energy region versus the low-energy region where the data is actually providing more meaningful constraints. The paper flags that weakness clearly.
Kai: And looking at their overall findings, they conclude that information-theoretic approaches offer a way to analyze these lattice correlators by quantifying those constraints through conditional KL divergence and joint constraints involving correlations across all energies.
Paper summary: Mira: The authors emphasize that this framework allows them to examine both the marginal updates at each energy and the joint constraints involving different energies, which helps identify which parts of the spectrum are constrained by the data and which remain largely determined by the prior.
Lev: It’s a useful tool for assessing reliability across different energy regions, but we have to remember that their method relies on defining those priors correctly in the first place. That's where most of our systematic uncertainty will likely land when we try to implement this on actual physics problems.
Kai: So to wrap up the core message of "An Information-Theoretic Perspective on Spectral Reconstruction in Lattice QCD," it’s that information theory provides a natural language for understanding how data updates assumptions, allowing us to see exactly where the data is pushing our reconstruction and where our initial model still dominates.
Mira: The paper's title really captures that focus on perspective; it’s not just about getting a better number, but understanding the fundamental role of the constraints supplied by the data relative to our prior knowledge.
Kai: So, what does this all mean for where we are now in quantum hardware and theory? We’ve seen how these information measures quantify the data's influence—the conditional K telling us about single energy updates and J telling us about the correlations across the whole spectrum.
Mira: It means that for spectral reconstruction, knowing where you are on that spectrum is actually tied to how much data you have at different energy scales. The paper’s title suggests this approach isn't just a technical trick but a fundamental way to understand the relationship between what we assume and what the noisy physical data tells us.
Lev: From my side, it’s about knowing where our real hardware runs best; if we can map out these constraint regions, we know which energy windows are reliable for extracting features. The main implication is that we need to be much more careful about selecting those priors because the posterior distribution is really governed by the prior in many places.
Kai: And that’s what they point toward for future work—using these information measures to guide our choice of priors and quantify those systematic uncertainties that come from picking a wrong starting assumption.
Mira: So, the authors are suggesting that this information-theoretic view gives us a systematic way to move beyond just getting one reconstructed spectrum and instead characterize the whole picture of constraints imposed by the experimental data on the theoretical model.
Conclusion: Kai: So, we're wrapping up this look at "An Information-Theoretic Perspective on Spectral Reconstruction in Lattice QCD." Basically, they're using information theory to figure out how much useful data you get from noisy lattice correlators when you try to reconstruct those spectral functions.
Mira: Yeah, the authors are showing how using Kullback-Leibler divergence lets them measure that exact info supplied by the data compared to their initial guesses. It’s not just a simple fit anymore; it’s about quantifying uncertainty directly from the input.
Lev: From an error correction standpoint, that means we can finally look at *why* our reconstructions fail or succeed based on how much data we actually fed in, rather than just guessing where the data ends up.
Kai: Exactly. They focus on separating the information update for a single energy value from the constraints you get across all the other energies simultaneously. That’s key for understanding how coupled those physical measurements are.
Mira: And they show that this framework lets them look at correlations, not just individual points, which is where most of our physical insight comes from in these systems. The whole point of this approach is to move beyond just picking a single best spectrum and instead characterizing the uncertainty structure itself.
Lev: If you take that to real hardware, it means when we’re calibrating our algorithms, we can use these information measures to see exactly which energy regions are actually providing meaningful data versus where the constraints are just noise from the correlator setup.
Kai: It changes how we think about model building because now we have a metric for data quality that goes beyond simple error bars. It’s not just about how much noise is there; it's about what information that noise actually carries.
Mira: And the authors are pointing toward using these measures to guide our choice of priors in the future, which is huge because picking a bad starting assumption can totally skew your final result, no matter how good your reconstruction algorithm is.
Lev: So they’re suggesting that instead of just optimizing one spectrum, we need a way to systematically quantify the systematic uncertainty that comes from choosing our initial model parameters. That’s where the real work lands.
Nan Wang, Long-cheng Gui, Jian Liang, Jun Shi, Han-yang Xing
State Key Laboratory of Nuclear Physics and Technology, Institute of Quantum Matter, South China Normal University · Key Laboratory of Atomic and Subatomic Structure and Quantum Control (MOE), Guangdong-Hong Kong Joint Laboratory of Quantum Matter, Guangzhou 510006, China · Guangdong Basic Research Center of Excellence for Structure and Fundamental Interactions of Matter, Guangdong Provincial Key Laboratory of Nuclear Science, Guangzhou 510006, China · Department of Physics, Hunan Normal University · Synergetic Innovation Center for Quantum Effects and Applications (SICQEA), Changsha 410081, China · Key Laboratory of Low-Dimensional Quantum Structures and Quantum Control Ministry of Education, Changsha 410081, China · Guangdong Provincial Key Laboratory of Sensing Physics and System Integration Applications, School of Physics and Optoelectronic Engineering, Guangdong University of Technology
hep-lat
Submitted: 2026-11-07
Updated: 2026-11-07
Comments: 9 pages, 6 figures
License: http://creativecommons.org/licenses/by-nc-nd/4.0/
Importance score: 86/100
The gist: Reconstructing spectral functions from finite, noisy lattice correlators requires understanding the constraints supplied by the data relative to prior assumptions This information-theoretic framework…
Key concepts
- Kullback–Leibler (KL) Divergence
- This measures the information gained by comparing two probability distributions: the prior distribution (what we assume before seeing data) and the posterior distribution (what we know after seeing the data). A higher KL divergence means more information is supplied by the new data, effectively narrowing down our uncertainty about a specific spectral value.
- Conditional Information Measure (K)
- This measure quantifies how much information the data provides about a single spectral value at one energy ($\omega_i$) compared to what we already knew from other energies. It tells us the reduction in uncertainty about that specific energy level when considering all other known spectral values.
- Full-Conditional Information Diagnostic (J)
- This advanced measure accounts for correlations across the entire spectrum. It quantifies the average reduction in uncertainty about one spectral value ($\rho_i$) when all other spectral values ($\rho_{-i}$) are known, while holding the overall prior assumption constant. It captures how knowing other energies helps constrain a specific energy level.
Terminology
Summary
Reconstructing spectral functions from finite, noisy lattice correlators requires understanding the constraints supplied by the data relative to prior assumptions This information-theoretic framework uses Kullback–Leibler (KL) divergences to quantify the information supplied by the data through changes from prior to posterior probability distributions The analysis accounts for both changes in the probability distributions of spectral strength at individual energies and constraints encoded in correlations across the spectrum
Information-Theoretic Framework
The framework compares the posterior density p with the prior density q using the KL divergence DKL(p∥q) = Z dx p(x) log2 p(x) q(x) This divergence measures the information supplied by the data relative to the prior The joint posterior of ρ and α is defined as p(ρ, α D) = p(D ρ) p(ρ α) p(α) The KL chain rule separates the information supplied by the data into an update of α and a spectral update at fixed α The energy-dependent measure K, with values Ki at ωi, measures the information supplied by the data about the spectral value at each energy relative to its conditional prior To include correlations, a full-conditional KL diagnostic J is introduced where Ji ≡ J(ωi) = Eα,ρ−iD h DKL p(ρi D, α, ρ−i)∥p(ρi α) This measure quantifies the average reduction in posterior uncertainty about ρi when the spectral values at all other energies are known, at fixed α
Bayesian Reconstruction and Numerical Evaluation
The reconstruction setup involves approximating the integral over a finite energy interval using Kti = ∆ωie−ωit, with ∆ωi the integration weight associated with ωi The spectral prior is defined by the BR functional T(ρ) = X i ∆ωi ρi mi− 1 − ln ρi mi, where mi = m(ωi) > 0 is the default model Posterior sampling in PyMC provides samples of both ρ and α while preserving their correlations The marginal posterior required for K is obtained by integrating over the remaining spectral values p(ρi D, α) = Z dρ−i p(ρi D, α, ρ−i) × p(ρ−i D, α) The information measures K and J are computed by averaging the KL divergence of each normalized conditional density in Eq. (15) relative to its corresponding prior p(ρi α) over the joint posterior samples
Mock-Data Results
The mock spectrum contains three equal-weight peaks, ρtrue(ω) = X 3 n=1 δ(ω − En), (E1, E2, E3) = (1, 1.5, 3) The reconstruction captures the main spectral structures with centroids approximately 1. p. 4 The corresponding peak areas in the posterior mean are 1. The reconstruction
captures the main spectral structures. The choice of a spectral estimate depends on
the properties one aims to recover [31]. In this mock test,
the MAP provides higher spectral resolution and reproduces the input peak positions and areas more faithfully
than the posterior mean. 00, Ki ≃ 109. 88 bits and
Ji ≃ 117.Where the paper says it
Where the paper says it
Where the paper says it
Lattice-Correlator Results
The analysis was applied to a proton two-point correlator calculated on the RBC/UKQCD 48I ensemble The information measures for the proton reconstruction show that K exhibits pronounced local dips near the two low-energy peaks The smoothed information curves provide a semiquantitative basis for assessing reconstruction reliability across energy regions In the high-energy region ω > 6 GeV, both smoothed measures decrease overall, indicating progressively weaker data constraints at higher energies
Summary and Outlook
Information-theoretic approaches offer a promising direction for analyzing lattice correlators and reconstructing spectral functions The conditional KL K quantifies the information supplied by the data through the change from the conditional prior to the posterior distribution of each spectral value The full-conditional KL J measures this information while also accounting for correlations with the rest of the spectrum Building on this study, future work will focus on using these developments to guide prior selection and quantify systematic uncertainties associated with prior choices
--- Page 4 ---
A variety of methods have been applied or developed to address this inverse problem, including the maximum entropy method (MEM) [14], Bayesian reconstruction (BR) [15], and the Backus-Gilbert (BG) method [13, 16] These approaches have been complemented by the Hansen-Lupo-Tantalo (HLT) method [17] and Chebyshev expansions [18], as well as Gaussianprocess regression [19, 20], sparse modeling [21], and neural networks [22–24] Further developments include analytic formulations using conformal maps [25] and Mellin transforms [26], together with convex optimization for bounding spectral integrals [27] Posterior sampling has also enabled Bayesian analyses beyond a single optimized spectrum [28]
--- Page 5 ---
The posterior distribution allows us to characterize the reconstructed spectrum through its mean, credible intervals and correlations Alongside the posterior mean, we consider the whole-spectrum maximum a posteriori (MAP) estimate, ρMAP = arg max ρ>0 p(ρ D), (14) We compare the posterior mean and MAP spectra in the numerical results
--- Page 8 ---
The local dips in K reflect its definition as a difference between posterior and prior distributions On either side of a peak, the reconstructed spectral strength changes between the small values in a valley and the large values near the peak center This suggests that the posterior distribution may also locally approach its conditional prior, providing a possible explanation for the dip in K For example, as shown in Fig. 3, between the valley to the left of the first peak and its center, a dip occurs at ω ≃ 0. p. 5 For example, as shown in
Fig. 3, between the valley to the left of the first peak and
its center, a dip occurs at ω ≃ 0. 95, where Ki ≃ 0. Its proximity to the crossing of the posterior and
prior means in Fig.Where the paper says it
Where the paper says it
--- Page 9 ---
The MAP selects the most probable spectrum, whereas the posterior mean averages over the full posterior and is sensitive to its asymmetry and extended tails The choice of a spectral estimate depends on the properties one aims to recover [31] In this mock test, the MAP provides higher spectral resolution and reproduces the input peak positions and areas more faithfully than the posterior mean Although the mean reflects additional aspects of the posterior distribution, the MAP is the better choice for representing narrow spectral structures studied here
--- Page 9 ---
The upper panel of Fig. 2 illustrates this limit: the highenergy posterior marginal at ω ≃ 7. p. 4 (22)
The upper panel of Fig. 2 illustrates this limit: the highenergy posterior marginal at ω ≃ 7.Where the paper says it
--- Page 9 ---
The posterior is therefore governed by the prior, so its mean approaches the weighted prior mean and its MAP approaches the prior mode m This high-energy behavior is consistent with the weak-data limit discussed in Eq. (22)
--- Page 9 ---
The information measures are overall smaller and show less pronounced peak structures, consistent with the reduced precision of the correlator data In the upper panel of Fig. 5, K exhibits pronounced local dips near the two low-energy peaks To understand the location of these dips, we turn to Fig. p. 6 In the upper panel of Fig. 5, K exhibits pronounced
local dips near the two low-energy peaks.Where the paper says it
--- Page 9 ---
At the two peak energies, K exhibits pronounced local dips near the two low-energy peaks To assess the constraints across each peak region while averaging over these localized variations, we use the Gaussiansmoothed measures in the lower panel At the two peak energies, K ≃ 3.47 and 1.51 bits, with J ≃ 5. p. 1 INFORMATION-THEORETIC
FRAMEWORK
To assess how strongly the data constrain different
parts of a reconstructed spectrum, we compare the posterior density p with the prior density q using the KL
divergence,
DKL(p∥q) = Z
dx p(x) log2
p(x)
q(x)
. (1)
We use logarithms to base two, so all information quantities are in bits. These values indicate that, across the
peak regions, the data appreciably modify the individual distributions relative to their conditional priors, with
additional constraints encoded in correlations. In the suppressed spectral region 2 ≲ ω ≲ 6 GeV, K
ranges from approximately 2. INFORMATION-THEORETIC
FRAMEWORK
To assess how strongly the data constrain different
parts of a reconstructed spectrum, we compare the posterior density p with the prior density q using the KL
divergence,
DKL(p∥q) = Z
dx p(x) log2
p(x)
q(x)
. (1)
We use logarithms to base two, so all information quantities are in bits.Where the paper says it
Where the paper says it
Where the paper says it
--- Page 9 ---
Using 1 bit as an illustrative reference, J suggests appreciable data constraints up to ω ≃ 4, consistent with the reconstructed spectrum in Fig.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Information Theoretic Constraint Quantification: The AI system can quantify how strongly reconstructed spectral structures are constrained by data and how reliable they are in light of prior effects by calculating conditional KL divergences K and J.
-
Joint Constraint Diagnosis: The system can distinguish between changes in individual marginal distributions at each energy and constraints encoded in correlations across the spectrum using the full-conditional KL diagnostic J, which is decomposed into a marginal KL divergence and a posterior mutual information contribution.
-
Energy Scale Smoothing for Regional Assessment: The AI system can obtain a smoothed assessment of information measures at a specified energy scale by applying Gaussian weights to neighboring energies, as described by the formula X(omega i) = P j delta omega j e -(omega i - omega j) squared / (2 sigma 2) Xj.
-
Reliability Assessment Criterion: The system can provide a semiquantitative criterion for assessing how strongly reconstructed spectral structures are constrained by the data and how reliable they are in light of prior effects, based on comparing the information measures K and J against a reference value like 1 bit.
-
Prior Influence Mapping: The AI system can determine regimes where the posterior distribution is governed by the prior, as shown when
the posterior distribution may also locally approach its conditional prior
near peak centers, which corresponds toa dip occurs at omega approximately 0.95, where Ki approximately 0.05 bits.
Abstract
Assessing the reliability of spectral reconstruction from finite, noisy lattice correlators requires understanding the constraints supplied by the data relative to prior assumptions. We develop an information-theoretic framework using Kullback--Leibler (KL) divergences to quantify the information supplied by the data through changes from prior to posterior probability distributions. Our analysis accounts for both changes in the probability distributions of spectral strength at individual energies and constraints encoded in correlations across the spectrum. We smear these information measures over neighboring energies using Gaussian weights to obtain a smoothed assessment at a specified energy scale. We demonstrate this framework using a hierarchical Bayesian reconstruction method with posterior sampling, applied to mock correlator data generated from a three-peak spectrum and a near-physical lattice proton correlator. These calculations show that the information measures help explain the properties of reconstructed spectra. They also provide a semiquantitative criterion for assessing how strongly reconstructed spectral structures are constrained by the data and how reliable they are in light of prior effects.
Sources
- Transport Properties of the Quark-Gluon Plasma -- A Lattice QCD Perspective
- Lattice QCD estimates of thermal photon production from the QGP
- Probing the energy-smeared R-ratio on the lattice
- Towards the nucleon hadronic tensor from lattice QCD
- Elastic and resonance structures of the nucleon from hadronic tensor in lattice QCD: implications for neutrino-nucleon scattering and hadron physics
- Inclusive hadronic decay rate of the $\tau$ lepton from lattice QCD: the $\bar u s$ flavour channel and the Cabibbo angle
- Inclusive semileptonic decays of the $D_{s}$ meson: Lattice QCD confronts experiments
- Inclusive semileptonic decays of the $D_s$ meson: A first-principles lattice QCD calculation
- Inclusive semileptonic $D_s\to X_s\ell\bar\nu$ decays from lattice QCD: continuum and chiral extrapolation
- Spectral-function determination of complex electroweak amplitudes with lattice QCD
- Theoretical framework for lattice QCD computations of $B\to K \ell^+ \ell^-$ and $\bar{B}_s\to \ell^+\ell^- \gamma$ decays rates, including contributions from charming penguin diagrams
- From deep inelastic scattering to heavy-flavor semi-leptonic decays: Total rates into multi-hadron final states from lattice QCD
- Maximum Entropy Analysis of the Spectral Functions in Lattice QCD
- A novel Bayesian approach to spectral function reconstruction
- On the extraction of spectral densities from lattice correlators
- Reconstruction of smeared spectral function from Euclidean correlation functions
- Reconstructing QCD Spectral Functions with Gaussian Processes
- Bayesian solution to the inverse problem and its relation to Backus-Gilbert methods
- Sparse modeling approach to obtaining the shear viscosity from smeared correlation functions
Related papers
- Heavy-Meson Bag Parameters using Gradient Flow
- Impact of Irrep Truncations on Quantum Simulations of QCD
- Evaluating generative samplers for lattice field theory by scale
- Renormalized symmetry probes for emergent symmetries and non-chiral lattice fermions
- Scalable Generative Sampling and Multilevel Estimation for Lattice Field Theories Near Criticality
- Lattice QCD study of the K*(892) resonance at the physical point