⚛️ nuclear theory

A Maximum-Entropy Method for Zero-Skewness Valence GPDs Constrained by Nucleon Electromagnetic Form Factors

This paper presents a reduced-profile maximum-entropy method that constructs constrained zero-skewness valence-quark generalized parton distribution transverse profiles by integrating nucleon electromagnetic form factors and forward parton distribution functions to establish a stable baseline for analyzing impact-parameter distributions.

Original authors: Seung-il Nam

Published 2026-07-03

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Seung-il Nam

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the proton (a core part of a nucleus) not as a solid marble, but as a bustling city made of tiny, fast-moving particles called quarks. Physicists have two main ways of looking at this city, but each method only shows them half the picture.

The Two Half-Pictures

The "Speed" Map (PDFs): One method tells us how much "longitudinal" momentum each quark has. It's like knowing how fast every car is driving down a highway, but it doesn't tell you where on the highway they are driving side-to-side.
The "Shape" Map (Form Factors): Another method tells us the overall shape and size of the proton when it gets hit by electricity. It's like seeing the shadow of the city from far away. You know the city's total outline, but you can't see the individual buildings or where the traffic is concentrated.

The Missing Puzzle Piece
The goal of this paper is to combine these two half-pictures into one complete 3D map. This new map, called a Generalized Parton Distribution (GPD), would show us not just how fast a quark is moving, but exactly where it is located inside the proton.

However, there is a huge problem: The "Shape" data (Form Factors) is like a blurry photo. It tells us the average position of all the cars, but it doesn't tell us where the fast cars are versus the slow cars. Mathematically, there are infinite ways to arrange the cars to create that same blurry shadow. This is called an "inverse problem," and it's notoriously difficult to solve because there are too many possible answers.

The Solution: The "Maximum Entropy" Chef
To solve this, the author, Seung-il Nam, uses a mathematical technique called the Maximum Entropy Method (MEM).

Think of it like a chef trying to recreate a secret soup recipe based only on the taste of the final broth.

The chef knows the ingredients (the "Forward PDFs" and "Form Factors").
The chef knows the final flavor (the "Constraints").
But the chef doesn't know the exact amounts of spices used at every step.

If the chef just guessed random amounts, the soup might taste right, but it would be chaotic and weird. The Maximum Entropy rule is the chef's golden rule: "Make the soup as smooth and simple as possible, adding no extra flavors or weird textures unless the taste data forces you to."

In this paper, the "chef" is a computer algorithm. It looks at the known data (the proton's shape and speed limits) and asks: "What is the smoothest, most logical arrangement of quarks that fits all the rules?"

The "Reduced Profile" Trick
The paper introduces a clever shortcut to make the math work. Instead of trying to guess the location of every single quark (which would be like guessing the position of every grain of sand on a beach), the author assumes the quarks follow a specific, simple pattern.

They imagine the quarks are arranged in layers. The "profile" is a rule that says: "As a quark gets faster (carries more momentum), it tends to huddle closer to the center of the proton."

The author uses a simple mathematical formula (a "reduced profile") to describe this huddling. This formula has only a few knobs to turn (parameters). By turning these few knobs, the computer can find a solution that fits the blurry "Shape" data perfectly without inventing fake, chaotic details.

What They Found
Using this method, the author successfully built a 3D map of the proton's interior:

The Result: They confirmed that fast-moving quarks (those with high momentum) are indeed more tightly packed in the center of the proton, while slower quarks are spread out more widely.
The Check: They tested their map against real-world data from particle accelerators (specifically looking at how light scatters off protons). The map predicted the scattering patterns correctly, proving the method works.
The Stability: They showed that even if they tweaked the "smoothness" rules slightly, the main features of the map stayed the same. This means the result isn't just a fluke; it's a robust discovery.

In Summary
This paper doesn't discover a new particle or change the laws of physics. Instead, it provides a new, reliable blueprint for reconstructing the 3D interior of the proton. It takes the blurry, incomplete data we have and uses a "smoothness" rule to fill in the gaps, giving us our clearest view yet of how quarks are arranged inside the matter that makes up our world.

Technical Summary: A Maximum-Entropy Method for Zero-Skewness Valence GPDs Constrained by Nucleon Electromagnetic Form Factors

Problem Statement
The spatial and momentum structure of the nucleon is described by two complementary but incomplete sets of observables: elastic electromagnetic form factors (FFs) and parton distribution functions (PDFs). FFs encode the transverse spatial distribution of the nucleon but integrate over all parton momentum fractions ( $x$ ), while PDFs describe the longitudinal momentum distribution but lack transverse spatial resolution. Generalized Parton Distributions (GPDs) unify these descriptions, yet reconstructing the $x$ -dependent transverse structure of valence quarks from elastic FFs alone constitutes a severely underdetermined inverse problem. Specifically, a single FF constrains only the $x$ -integrated moments of the GPDs, leaving the local $x$ -dependent transverse profile arbitrary. The challenge is to construct a physically motivated, stable, and least-biased zero-skewness valence GPD framework that satisfies four nucleon FF constraints ( $F_1^p, F_1^n, F_2^p, F_2^n$ ) and forward-limit normalizations without introducing unphysical oscillations or spurious structure.

Methodology
The authors formulate a reduced-profile Maximum-Entropy Method (MEM) framework to solve this inverse problem. The approach relies on the following key components:

Ansatz and Constraints: The zero-skewness valence GPDs, $H_v^q(x,t)$ and $E_v^q(x,t)$ , are modeled using an exponential $t$ -profile ansatz:
$H_v^q(x, t) = q_v(x) \exp[t f_H^q(x)], \quad E_v^q(x, t) = e_v^q(x) \exp[t f_E^q(x)]$
Here, $q_v(x)$ and $e_v^q(x)$ are the forward-limit valence PDFs and Pauli-type forward distributions, respectively. The unknown functions $f_H^q(x)$ and $f_E^q(x)$ represent the positive transverse profile functions. The constraints include:
- Elastic Moments: The first Mellin moments of the GPDs must reproduce the flavor-separated Dirac and Pauli form factors derived from experimental proton and neutron FFs.
- Forward Normalizations: At $t=0$ , the GPDs must reduce to the input PDFs ( $q_v$ ) and satisfy flavor anomalous magnetic moment normalizations ( $\kappa^q$ ).
Reduced Profile Manifold: To regularize the underdetermined inverse problem, the authors avoid a direct grid-based inversion of the profile functions. Instead, they restrict the profiles to a low-dimensional, positive manifold using a specific functional form:
$f(x) = 0.05 + (1-x)^2 \exp(c_0 + c_1 x + c_2 x^2)$
This reduces the degrees of freedom from a large grid of points to 12 shape parameters (three coefficients for each of the four profiles: $f_H^u, f_H^d, f_E^u, f_E^d$ ). This reduction suppresses local modes that elastic moments cannot constrain.
Entropy Regularization: The MEM functional $Q = \alpha S - \frac{1}{2}\chi^2$ is maximized. The likelihood term $\chi^2$ measures the deviation between the reconstructed FFs and the input data. The entropy term $S$ (Shannon-Jaynes) penalizes deviations from a smooth, positive default model, ensuring the solution is the "least-biased" profile compatible with the constraints. The hyperparameter $\alpha$ balances data fidelity against prior smoothness.
Numerical Implementation: The calculation uses analytic beta-function parametrizations for the forward inputs and a smooth dipole-based parametrization for the elastic FFs as a proof-of-concept. The optimization is performed using the L-BFGS-B algorithm on a discrete grid, with forward inputs normalized on the same grid to ensure exact machine-precision satisfaction of sum rules at $t=0$ .

Key Results

Reconstruction Stability: The reduced-profile solution is highly stable under variations of the entropy weight $\alpha$ (tested in the range $10^{-4}$ to $10^{-2}$ ). The maximum pointwise deviation in the profile functions remains below 0.01%, indicating that the solution is driven by the form-factor constraints rather than the entropy prior.
Form Factor Reproduction: The reconstructed GPDs reproduce the input Sachs form factors ( $G_E^p, G_M^p, G_E^n, G_M^n$ ) at the percent level across the range $0 \le Q^2 \le 5 \text{ GeV}^2$ . The forward normalizations for valence quark numbers and anomalous magnetic moments are satisfied exactly on the discrete grid.
Physical Interpretation: The resulting profiles $f^q(x)$ are positive and decrease as $x \to 1$ . This leads to impact-parameter distributions where large- $x$ partons are more localized near the transverse center of momentum (transverse shrinkage), consistent with physical expectations.
Qualitative Checks: A proxy for the beam-spin asymmetry ( $A_{LU}$ ) in exclusive photon electroproduction was constructed using the reconstructed profiles. The proxy successfully reproduces the leading $\sin \phi$ azimuthal pattern observed in CLAS12 data, providing a qualitative consistency check beyond the elastic constraints.

Significance and Claims
The paper claims to provide a controlled zero-skewness baseline for connecting elastic form-factor constraints to $x$ -dependent transverse profiles. The authors emphasize that the novelty lies not in the exponential $t$ -profile ansatz itself, which is standard, but in the formulation of the four-form-factor problem as a single coupled MEM pipeline. This pipeline integrates elastic moment constraints, forward-PDF normalizations, and anomalous-moment constraints within an entropy-based regularizer that preserves positivity.

The work is presented as a proof-of-concept demonstrating that a physically motivated, low-dimensional transverse profile can be extracted from elastic data and forward information without overfitting. The authors explicitly state that the results are conditional on the chosen reduced profile manifold and the specific analytic inputs used. They do not claim a global phenomenological extraction but rather a stable starting point for future analyses that can incorporate empirical form-factor fits with covariance information, modern PDF sets, lattice-QCD inputs, and hard exclusive observables. The method is designed to be input-agnostic, allowing the reconstruction machinery to be applied to different parametrizations without structural modification.

More like this