arXiv subjects

Nobuo Sato

Publications and source records attributed to Nobuo Sato.

At least 19 recordsLinked to original sources

arXiv2026

Learning Transverse Momentum Distributions from Raw Scattering Events via Conditional Diffusion

Extracting transverse momentum dependent parton distribution functions (TMD PDFs) from semi-inclusive deep inelastic scattering (SIDIS) data is a central goal of the nucleon structure program at Jefferson Lab and the future Electron-Ion Collider. Traditional extraction methods rely on parameterized functional forms and iterative fitting, which can limit the flexibility of the resulting distributions and make uncertainty quantification cumbersome. We present a conditional diffusion model that learns to map raw SIDIS event kinematics directly to TMD PDFs, bypassing explicit functional assumptions. Evaluated on simulated SIDIS data at CLAS12 kinematics, the model recovers the underlying TMD with informative uncertainties that narrow steadily with increasing event statistics, and produces reliable estimates even with as few as 1,000 conditioning events, a statistics-limited regime directly relevant to ongoing and planned experiments.

hep-ph↗

arXiv2026

Active Diffusion-Based Inference for Ill-Posed Inverse Problems under Incomplete Priors

Many scientific and engineering applications require estimating unknown parameters from experimentally observable data -- an inverse problem that is inherently challenging due to nonlinearity, noise, and ill-posedness. In this paper, we propose an active diffusion-based inverse problem solver. A DM is trained to learn the mapping between the parameter space and the observable space. By iteratively detecting and correcting model misspecification through posterior uncertainty, the method discovers and learns the correct region of parameter space, even when initial training bounds exclude the true parameters. This provides a principled, Bayesian justification for adaptive domain augmentation and ensures robust inference for inverse problems under incomplete prior knowledge. We demonstrate the effectiveness of our inverse solver for a toy inverse problem with infinite solutions, and for the parameterization of the quantum correlation functions to event observables in a Quantum Chromodynamics analysis of nucleon structure.

stat.ML↗

arXiv2026

Multi-Dataset Inverse Problem Solving with Distributed Generative AI

Extracting a shared set of unknown, not directly measurable quantities from multiple, heterogeneous datasets is a common challenge across scientific domains. A prominent example is the combination of datasets obtained from different measurements with different settings (e.g. varying detector resolutions). Analyzing such datasets jointly, rather than independently or after naive merging, is essential for obtaining precise and unbiased estimates of the unknowns, but requires careful treatment of dataset heterogeneity and is computationally demanding. We present a generalized framework for simultaneously analyzing multiple heterogeneous datasets in the context of generative AI-based inverse problem solvers. Building on our recent Scalable Asynchronous Generative Inverse Problem Solver (SAGIPS) framework, we extend the well-established distributed data-parallel training paradigm to non-identically distributed datasets, where each dataset is controlled by the same set of unknown inference parameters but covers a different region of the available feature space. Each dataset is processed through its own forward operator and discriminator, providing complementary constraints that collectively guide a shared generator toward global parameter consistency. We validate the approach using a controlled setup inspired by a multi-detector scattering experiment. We provide numerical evidence that our framework is robust to different data fidelities, which arise from unknown detector systematics in the Rutherford experiment, and we show the scaling behavior on multi-GPU leadership computing systems. The results show that our approach is well suited for real-world multi-dataset analyses in which experimental conditions vary across measurements.

cs.DC↗

arXiv2026

Impact of Future Dihadron Production Measurements on the Transversity Distributions and Tensor Charges of the Nucleon

We assess the impact of future measurements of dihadron production in semi-inclusive deep-inelastic scattering from the CLAS12 and proposed SoLID experiments at Jefferson Lab, as well as from the ePIC experiment at the future Electron-Ion Collider (EIC), on the transversity parton distribution functions (PDFs) and the corresponding tensor charges of the nucleon. To this end, we generate pseudo-data for these experiments for a proton target (CLAS12 and ePIC) and a $^3$He target (SoLID and ePIC), and we include these pseudo-data in the JAMDiFF global analysis of existing experimental dihadron data. We find that future data from Jefferson Lab will significantly reduce uncertainties in the transversity PDFs in the region of intermediate-to-large quark momentum fractions $x$, while the EIC will provide strong constraints across the entire range of $x$, allowing for the first experimental test of the predicted small-$x$ behavior of the transversity PDFs. In discussing the reduction of uncertainties in the tensor charges, we also compare the results from the data analyses with those from lattice QCD, highlighting scenarios in which compatibility or tension between the two would arise.

hep-ph↗

arXiv2026

An explicit Galois descent for multiple $t$-values of maximal height

We give an explicit formula for the Galois descent expressing multiple $t$-values of maximal height in terms of classical multiple zeta values, making precise Murakami's earlier motivic result. Our results rely on the theory of iterated beta integrals. We apply this formula to obtain evaluations of various multiple zeta-half values.

math.NT↗

arXiv2026

TMDs in the Lens of Generative AI: A Pixel-Based Approach to Partonic Imaging

This work introduces a novel, nonparametric pixel-based framework for the Bayesian inference and imaging of transverse momentum dependent (TMD) parton distributions. The methodology is built upon a fully differentiable framework that integrates TMD evolution with the Collins-Soper-Sterman formalism, enabling the simultaneous extraction of partonic distributions and the nonperturbative evolution kernel. To achieve efficient and exact sampling of the high-dimensional posterior, we leverage generative AI through a hybrid normalizing flow-driven Metropolis-Hastings approach. The framework is validated through multi-scale closure tests of increasing complexity, ranging from basic functional models to convoluted structure functions. Using singular value decomposition (SVD), we rigorously characterize the uncertainty of the reconstructed distributions and reveal the existence of null TMDs, which are functional components in the null space of the kernel that remain unconstrained by observables. The new framework provides the first integration of pixel-based discretization, generative AI, and SVD within a Bayesian context to solve the TMD inverse problem. This synergy between machine learning and multi-scale data removes inherent degeneracies and enables unbiased 3D partonic imaging.

hep-ph↗

arXiv2026

Iterated beta integrals

We introduce iterated beta integrals, a new class of iterated integrals on the universal abelian covering of the punctured projective line that unifies hyperlogarithms and classical beta integrals while preserving their fundamental properties. We establish various analytic properties of these integrals with respect to both the exponent parameters and the main variables. Their key feature is invariance under simultaneous translation of the exponent parameters, which generates relations between integrals over possibly different coverings. This mechanism recovers notable identities for multiple zeta values and variants -- including Zagier's 2-3-2 formula, Murakami's $t$-value analogue, Charlton's $t$-value analogue, Zhao's $2$-$1$ formula, and Ohno's relation -- and also yields new relations, such as a proof of a Galois descent phenomenon for multiple omega values.

math.NT↗

arXiv2026

Algebraic Quantum Intelligence: A New Framework for Reproducible Machine Creativity

Large language models (LLMs) have achieved remarkable success in generating fluent and contextually appropriate text; however, their capacity to produce genuinely creative outputs remains limited. This paper posits that this limitation arises from a structural property of contemporary LLMs: when provided with rich context, the space of future generations becomes strongly constrained, and the generation process is effectively governed by near-deterministic dynamics. Recent approaches such as test-time scaling and context adaptation improve performance but do not fundamentally alter this constraint. To address this issue, we propose Algebraic Quantum Intelligence (AQI) as a computational framework that enables systematic expansion of semantic space. AQI is formulated as a noncommutative algebraic structure inspired by quantum theory, allowing properties such as order dependence, interference, and uncertainty to be implemented in a controlled and designable manner. Semantic states are represented as vectors in a Hilbert space, and their evolution is governed by C-values computed from noncommutative operators, thereby ensuring the coexistence and expansion of multiple future semantic possibilities. In this study, we implement AQI by extending a transformer-based LLM with more than 600 specialized operators. We evaluate the resulting system on creative reasoning benchmarks spanning ten domains under an LLM-as-a-judge protocol. The results show that AQI consistently outperforms strong baseline models, yielding statistically significant improvements and reduced cross-domain variance. These findings demonstrate that noncommutative algebraic dynamics can serve as a practical and reproducible foundation for machine creativity. Notably, this architecture has already been deployed in real-world enterprise environments.

cs.AI↗

arXiv2025

New framework for extracting GPDs from exclusive photon electroproduction

Recently, a new framework for studying generic $2 \to 3$ hard exclusive reactions, referred to as single-diffractive hard exclusive processes, has been introduced to provide a cleaner separation of the underlying physical mechanisms. In this work, we expand this formalism to the case of exclusive real-photon electroproduction off a nucleon, $e(\ell) + N(p) \to e(\ell') + N(p') + \gamma(q')$, which represents the classical channel for accessing generalized parton distributions (GPDs) in nucleons and nuclei. This extension enables a more systematic and physically transparent formulation of the reaction dynamics, paving the way for improved extractions of GPDs from experimental data as compared to existing approaches.

hep-ph↗

arXiv2025

Variational Neural Network Approach to QFT in the Field Basis

We present a variational neural network approach for solving quantum field theories in the field basis, focusing on the free Klein-Gordon model formulated in momentum space. While recent studies have explored neural-network-based variational methods for scalar field theory in position space, a systematic benchmark of the analytically solvable Klein-Gordon ground state -- particularly in the momentum-space field basis -- has been lacking. In this work, we represent the ground-state wavefunctional as a neural network defined on a discretized set of field configurations and train it by minimizing the Hamiltonian expectation value. This framework enables direct comparison to exact analytic results for a range of key observables, including the ground-state energy, two-point correlators, expectation value of the field, and the structure of the learned wavefunctional itself. Our results provide quantitative diagnostics of accuracy and demonstrate the suitability of momentum space for benchmarking neural network approaches, while establishing a foundation for future extensions to interacting models and position-space formulations.

hep-ph↗

arXiv2025

Pionic gluons from global QCD analysis of experimental and lattice data

We perform the first global QCD analysis of parton distribution functions (PDFs) in the pion, with lattice-QCD data on gluonic pseudo--Ioffe-time distributions fitted simultaneously with experimental Drell-Yan and leading neutron electroproduction data. Inclusion of the lattice results with parametrized systematic corrections significantly reduces the uncertainties on the gluon PDF at parton momentum fractions $x \gtrsim 0.2$, revealing a higher gluon density in the pion at large $x$ than in the proton. The similar gluon momentum fractions in the pion and proton further suggests a relative suppression of the pion gluon density at small $x$.

hep-ph↗

arXiv2025

Toward an event-level analysis of hadron structure using differential programming

Reconstructing the internal properties of hadrons in terms of fundamental quark and gluon degrees of freedom is a central goal in nuclear and particle physics. This effort lies at the core of major experimental programs, such as the Jefferson Lab 12 GeV program and the upcoming Electron-Ion Collider. A primary challenge is the inherent inverse problem: converting large-scale observational data from collision events into the fundamental quantum correlation functions (QCFs) that characterize the microscopic structure of hadronic systems within the theory of QCD. Recent advances in scientific computing and machine learning have opened new avenues for addressing this challenge using deep learning techniques. A particularly promising direction is the integration of theoretical calculations and experimental simulations into a unified framework capable of reconstructing QCFs directly from event-level information. In this work, we introduce a differential sampling method called the local orthogonal inverse transform sampling (LOITS) algorithm. We validate its performance through a closure test, demonstrating the accurate reconstruction of a test distribution from sampled events using Generative Adversarial Networks. The LOITS algorithm provides a central building block for addressing inverse problems involving QCFs and enables end-to-end inference pipelines within the framework of differential programming.

hep-ph↗

arXiv2025

First study of polarized proton-proton scattering with small-$x$ helicity evolution

We perform a phenomenological study of helicity-dependent parton distribution functions (PDFs) using small-$x$ helicity evolution equations, incorporating for the first time single-inclusive jet production data in polarized proton-proton ($pp$) scattering at parton momentum fractions $x < 0.1$. We also simultaneously include double-longitudinal spin asymmetries in inclusive and semi-inclusive deep-inelastic scattering probing $x < 0.1$. Employing the polarized small-$x$ pure-glue calculation of $pp\to gX$ for the jet production cross section, we modify the large-$N_c\&N_f$ KPS-CTT evolution equations by setting $N_f = 0$ to replicate the large-$N_c$ (pure-glue) limit, while retaining external quark flavors for the spinor field operators. We find that the $pp$ data have a considerable impact on the helicity PDFs at small $x$, reducing their uncertainties and leading to a total quark and gluon helicity in the proton for $x < 0.1$ of $-0.04 \pm 0.23$. Combining our analysis with the a recent JAM helicity PDF analysis of the world polarized data, which includes $x > 0.1$, we find a total quark and gluon helicity contribution for $x > 10^{-7}$ of between 0.02 and 0.51.

hep-ph↗

arXiv2024

Point cloud-based diffusion models for the Electron-Ion Collider

At high-energy collider experiments, generative models can be used for a wide range of tasks, including fast detector simulations, unfolding, searches of physics beyond the Standard Model, and inference tasks. In particular, it has been demonstrated that score-based diffusion models can generate high-fidelity and accurate samples of jets or collider events. This work expands on previous generative models in three distinct ways. First, our model is trained to generate entire collider events, including all particle species with complete kinematic information. We quantify how well the model learns event-wide constraints such as the conservation of momentum and discrete quantum numbers. We focus on the events at the future Electron-Ion Collider, but we expect that our results can be extended to proton-proton and heavy-ion collisions. Second, previous generative models often relied on image-based techniques. The sparsity of the data can negatively affect the fidelity and sampling time of the model. We address these issues using point clouds and a novel architecture combining edge creation with transformer modules called Point Edge Transformers. Third, we adapt the foundation model OmniLearn, to generate full collider events. This approach may indicate a transition toward adapting and fine-tuning foundation models for downstream tasks instead of training new models from scratch.

hep-ph↗

arXiv2024

Unified and optimal frame choice for generalized parton distributions

Reconstructing the internal three-dimensional quark and gluon structures of hadrons through generalized parton distributions (GPDs) from hard exclusive scattering processes is one of the most challenging tasks in nuclear and particle physics. In this paper, we introduce a new optimized reference frame that, for the first time, enables a unified view of all the reactions sensitive to GPDs and facilitates the interpretation of a variety of phase-space patterns that were previously hardly accessible and interpretable. Similarly to how the heliocentric description advanced our understanding of the solar system and gravitation, our new frame centers around a quasireal state, allows for a consistent separation of physical scales, and reveals a novel quantum interference mechanism.

hep-ph↗

arXiv2024

A criterion for $p$-closedness of derivations in dimension two

Jacobson developed a counterpart of Galois theory for purely inseparable field extensions in positive characteristic. In his theory, a certain type of derivations replace the role of the generators of Galois groups. This article provides a convenient criterion for determining such derivations in dimension two. We also present examples demonstrating the efficiency of our criterion.

math.AG↗

arXiv2023

Diffusion model approach to simulating electron-proton scattering events

Generative AI is a fast-growing area of research offering various avenues for exploration in high-energy nuclear physics. In this work, we explore the use of generative models for simulating electron-proton collisions relevant to experiments like CEBAF and the future Electron-Ion Collider (EIC). These experiments play a critical role in advancing our understanding of nucleons and nuclei in terms of quark and gluon degrees of freedom. The use of generative models for simulating collider events faces several challenges such as the sparsity of the data, the presence of global or event-wide constraints, and steeply falling particle distributions. In this work, we focus on the implementation of diffusion models for the simulation of electron-proton scattering events at EIC energies. Our results demonstrate that diffusion models can accurately reproduce relevant observables such as momentum distributions and correlations of particles, momentum sum rules, and the leading electron kinematics, all of which are of particular interest in electron-proton collisions. Although the sampling process is relatively slow compared to other machine learning architectures, we find diffusion models can generate high-quality samples. We foresee various applications of our work including inference for nuclear structure, interpretable generative machine learning, and searches of physics beyond the Standard Model.

hep-ph↗

arXiv2023

Global analysis of polarized DIS & SIDIS data with improved small-$x$ helicity evolution

We analyze the world polarized deep-inelastic scattering (DIS) and semi-inclusive DIS (SIDIS) data at low values of $x < 0.1$, using small-$x$ evolution equations for the flavor singlet and nonsinglet helicity parton distribution functions (hPDFs). The hPDFs for quarks, antiquarks, and gluons are extracted and evolved to lower values of $x$ to make predictions for the future Electron-Ion Collider (EIC). We improve on our earlier work by employing the more realistic large-$N_c\, \& N_f$ limit of the revised small-$x$ helicity evolution, and incorporating running coupling corrections along with SIDIS data into the fit. We find an anti-correlation between the signs of the gluon and $C$-even quark hPDFs as well as the $g_1$ structure function. While the existing low-$x$ polarized DIS and SIDIS data are insufficient to constrain the initial conditions for the polarized dipole amplitudes in the helicity evolution equations, future EIC data will allow more precise predictions for hPDFs and the $g_1$ structure function for $x$ values beyond those probed at the EIC. Using the obtained hPDFs, we discuss the contributions to the proton spin from quark and gluon spins at small $x$.

hep-ph↗