Symmetry adapted tensor networks and quantum classical geometric learning for hadronic tomography: a proton construction
summary
The gist
We construct a calibrated model of the proton generalized transverse-momentum-dependent distribution (GTMD) F1,1 in the real…
In short
A calibrated model of a proton distribution (GTMD F1,1) was constructed using a symmetry-resolved tree tensor network architecture combining experimental data and lattice inputs. This method showed that forward transverse information significantly constrains the distribution, narrowing its possible range substantially. Furthermore, it revealed a specific deformation that shifts the nonforward value from 0.588 to 0.378 GeV−2 while leaving other criteria unchanged.
Key concepts
- GTMD F1,1
- This is a calibrated model describing the proton's generalized transverse-momentum-dependent distribution. It is used to understand how the proton's internal structure changes depending on both its momentum and how it interacts with other particles, specifically focusing on valence-u quarks in the real sector.
- Symmetry-resolved tree tensor network
- This is the mathematical architecture used to organize all experimental measurements, lattice calculations, and phenomenological distributions. It structures these inputs by considering symmetry constraints (like color singlet structure) and exchange statistics, allowing the model to systematically combine different types of data into a coherent physical representation.
- Deformation
- This refers to a specific mathematical change applied to the model that leaves standard forward and elastic criteria untouched. This deformation demonstrated that even small changes in the underlying physics—specifically affecting the nonforward value—can cause significant shifts in observable quantities, moving a value from 0.588 to 0.378 GeV−2.
- Classical and quantum geometric nodes
- These represent the residual transverse response within the model's architecture. They are used alongside hadronic representations and effective currents to capture complex physical interactions, allowing the model to distinguish between different aspects of the proton's structure based on symmetry and physical metrics.
Terminology used across episodes
This episode discusses
- Symmetry adapted tensor networks and quantum classical geometric learning for hadronic tomography: a proton construction · Paper Radio
- Generalized parton correlation functions for a spin-1/2 hadron
- Quark Wigner Distributions and Orbital Angular Momentum
- Proper definition and evolution of generalized transverse momentum distributions
- Matching generalised transverse-momentum-dependent distributions onto generalised parton distributions at one loop
- One-loop matching for leading-twist generalised transverse-momentum-dependent distributions
- Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
- Determination of unpolarized TMD distributions from the fit of Drell-Yan and SIDIS data at N 4 LL
- Combination of Measurements of Inclusive Deep Inelastic e plus or minusp Scattering Cross Sections and QCD Analysis of HERA Data
- Tensor network decompositions in the presence of a global symmetry
- Quantum Circuit Learning
- Quantum Natural Gradient
- A Higher Order Perturbative Parton Evolution Toolkit (HOPPET)
- Generalized parton distributions in the deuteron
- Generalized parton distributions in spin-3/2 particles
- Transversity generalized parton distributions in spin-3/2 particles
- The isovector axial form factor of the nucleon from lattice QCD
- Moments of nucleon generalized parton distributions from lattice QCD simulations at physical pion mass
- Flavor diagonal nucleon charges using clover fermions on MILC HISQ ensembles
- Measurements of electron-proton elastic cross sections for 0.4 < Q squared < 5.5 (GeV/c) squared
The paper
Symmetry adapted tensor networks and quantum classical geometric learning for hadronic tomography: a proton construction · Read on arXiv
Dustin Keller
University of Virginia
Transcript
Introduction to the show: ident: High Energy Physics: Phenomenology. The day in particle phenomenology: models, signatures and constraints.
Kai: I'm Kai.
Mira: And I'm Mira, here with Lev.
Kai: We're looking at the paper: "Symmetry adapted tensor networks and quantum classical geometric learning for hadronic tomography".
Mira: Here's the gist of the paper. We construct a calibrated model of the proton generalized transverse-momentum-dependent distribution (GTMD) F1,1 in the real, staple-even valence-u sector at zero skewness The simultaneous dependence on quark momentum and momentum transfer is untouched by any of them: a deformation that leaves every present input invariant moves the nonforward value from 0.588 to 0.378 GeV−2 Model Construction and Architecture The construction organizes experimental measurements, lattice matrix elements, and phenomenological distributions through a single shared light-front state, organized as a symmetry-resolved tree tensor network in which color-singlet structure, exchange statistics, and the additive Jz and Iz charges are built Classical and quantum geometric nodes represent the residual transverse response The architecture combines a finite, symmetry-resolved hadronic representation, shared effective currents, and classical and quantum geometric function blocks Symmetry determines admissible transformations, invariant features condition shared responses, and the physical state metric identifies redundant coordinates The tensor-tree construction exposes collective longitudinal (L), Jacobi-orbital (O), color-singlet multiplicity (C), and joint spin–flavor–helicity (S) registers as physical factor labels How the Training Signal is Built Each input enters the shared state through its own response map with its published covariance, and we keep track of which inputs rest on common underlying data The calibration gives two main results, the first is that forward transverse information matters a great deal At x = 0.4 and k = 0.2 GeV, requiring compatibility with the ART projections narrows the attainable range of F1,1 by 81% within the kernel family we use The second is that the joint dependence on quark transverse momentum and momentum transfer is a direction none of the present inputs reaches We write down a deformation, equal to one at ∆ = 0 and 1 + O(r2) at short distance, that leaves every forward and elastic criterion unchanged and moves the nonforward value from 0.588 to 0.378 GeV−2 Where the Gains Come From The construction is designed to extend to the neutron, the deuteron, and spin-3/2 nuclei where symmetry and exchange constraints grow quickly The proton construction reported here is the first validated step of that program The final fit uses a 114-dimensional carrier, validated subsequently, and a normalized native real state chart with 113 local tangent directions The forward transverse information narrows the GTMD substantially, at x = 0.4 and k = 0.2 GeV the reference forward value is 0.978GeV−2, with an attained inner finite-kernel range of
0.813, 0.997: GeV−2 The joint momentum–transfer dependence is left to the model, and we can say by how much Validation and Sensitivity Analysis The finite-window checks have a clear quantitative effect, at x = 0.4, the hard-4GeV impact map has a maximum of 1.536 fm−2GeV−2 The largest absolute changes over its displayed plane are 0.334, 0.068, and 0.023 fm−2GeV−2 for the hard-2GeV, hard-3GeV, and cosine-tapered windows, respectively The joint stress of Eq.
Kai: (27) gives a second example at the nonforward anchor it changes 0.588 to 0.378, retaining the forward value and all current input criteria The deformation is given explicitly, and its effect on the momentum-space and phase-space images is shown in Sec.
Mira: VI The second statement sets the agenda, a useful future constraint has to respond to a deformation like Eq.
Paper summary: Kai: So to recap, the paper "Symmetry adapted tensor networks and quantum classical geometric learning for hadronic tomography: a proton construction" is proposing a way to build a calibrated model of the generalized transverse-momentum-dependent distribution F1,1 for the proton <ref:2611.14711#pg1,Symmetry adapted tensor networks and quantum classical geometric learning for hadronic tomography>. They do this by organizing experimental data and lattice results into one shared light-front state using symmetry-resolved tree tensor networks.
Mira: The thesis is that this architecture handles color singlet structure, exchange statistics, and additive Jz and Iz charges simultaneously. This allows them to treat classical and quantum geometric nodes as representing the residual transverse response of the proton.
Lev: The methodology involves a shared light-front state where every input enters through its own response map, tracking which data points they share with others. They use soft subtraction, one-loop nonsinglet matching, and NLO collinear evolution to handle the different scales.
Kai: The results focus on calibration against correlated transverse-momentum projections from ART25 TMD extraction, HERA inclusive data, and lattice matrix elements to see what they can determine about the proton's structure.
Mira: One of the main findings is that forward transverse information is quite important. At x equals zero point four and k equals zero point two GeV, compatibility with ART projections narrows the attainable range of F1,1 by eighty-one percent within their kernel family <ref:2611.14711#pg1>.
Lev: And they also found that the joint dependence on quark transverse momentum and momentum transfer is a direction that none of the inputs reach directly. That's a key piece of information about where current measurements fall short.
Kai: The paper concludes this calibration by writing down a specific deformation, equal to one at zero and one plus O of r squared at short distance, which leaves forward and elastic criteria unchanged but moves the nonforward value from zero point five eight eight to zero point three seven eight GeV minus two.
Mira: This deformation is the concrete mathematical statement showing what the present data determine about the proton’s five-dimensional structure, and what kind of measurement would be needed to find out the rest. It links directly to their discussion on dense momentum-space surfaces and finite-window phase-space maps shown in section six.
Lev: It sets up a clear direction for future work, suggesting that a useful constraint in the future has to respond specifically to a deformation like that one they wrote down. That's what drives the next step for any theory trying to move this forward.
Conclusion: Kai: Looking at the whole paper "Symmetry adapted tensor networks and quantum classical geometric learning for hadronic tomography: a proton construction", it’s really about building a robust framework to handle proton structure using symmetry. They show how you can unify diverse inputs into one state, which is powerful.
Mira: What this means in simpler terms is that they’ve created a way to take all those different experimental results and lattice calculations and see how they fit together within this tensor network structure. It defines what the proton looks like based on those measurements.
Lev: For someone focused on real hardware, the implication is that if we want to run simulations of this kind, we need to be prepared for that complexity in managing the large symmetry intertwiners and ensuring our error correction can handle that scale.
Kai: The paper’s findings about the forward transverse information narrowing F1,1 substantially at x equals zero point four and k equals zero point two GeV gives a very concrete constraint on what we expect to measure in those specific regions <ref:2611.14711#pg1>.
Mira: It shows that this model is not just fitting data; it’s defining the geometry of the proton structure itself by how these inputs are constrained. It moves beyond simple parameter fitting into a more fundamental structural description.
Lev: The paper explicitly notes that their construction is designed to be extensible, moving from the proton to heavier nuclei where nuclear composition becomes a tensor product of those validated subunits. That’s where the real challenge for scaling up this approach lies.
Kai: So we've seen how they use symmetry adapted tensor networks to handle complex dependencies between momentum and transfer, and that it leads to concrete numerical constraints on things like the nonforward value of F1,1 <ref:2611.14711#pg1>. It’s a model built on structure rather than just curve fitting.
More episodes
- 2610.38463-Gauge-parameter cancellation and finite form factors in one-loop quark scattering in the refined Gribov-Zwanziger framework
- 2610.37198-Searches for new particle physics by methods of classical and neutrino astronomy
- 2511.15779-SURFing to the Fundamental Limit of Jet Tagging
- 2505.01492-Photo- and Hadrodisintegration constraints on massive relics decaying into neutrinos
- 2505.20389-Dispersion relation of the neutrino plasma: Unifying fast, slow, and collisional instabilities
- 2511.10731-Keeping It Renormalizable: Minimal Baryogenesis induced Asymmetric Dark Matter
- 2404.04021-Vacuum correlations of the stress-energy-momentum tensor with constituent quarks
- 2503.02931-Relativistic Navier-Stokes description of the quark-gluon plasma radial flow
- 2606.06105-Bulk viscosity of a binary mixture: the role of the intra-species interaction
- 2607.03343-First look at the evaluation of two-loop Feynman integrals for radiative return processes