Functional anatomy of Pythia-Herwig differences with Kolmogorov-Arnold networks

Arghya Chattopadhyay

hep-phAug 16, 2026

Functional anatomy of Pythia-Herwig differences with Kolmogorov-Arnold networks

Authors: Arghya Chattopadhyay

Organizations: Department of Physics, University of Puerto Rico at Mayagüez Mayagüez, Puerto Rico 00681, USA

Abstract

Differences between high-energy event generators can arise at several stages of the collision simulation, from the hard scattering through parton showering and hadronization to the final event. These differences are usually summarized using observable distributions or global classifier scores. While these quantify the disagreement, they do not reveal which observable-level structures carry it or whether those structures persist through different stages of event generation. In this work, we formulate this problem as a staged functional analysis of generator-model differences. Following the same hard dijet events through Pythia and Herwig at shower-only, hadronized, and full-generator levels, we use an additive Kolmogorov-Arnold network (KAN) representation of the classifier-derived log density ratio to decompose the learned discrepancy into explicit one-dimensional observable responses that can be isolated, recomposed, and transported between generator stages. Within the same eight-observable jet representation, the Pythia-Herwig difference is driven mainly by multiplicity at shower level, shifts toward jet mass and shape after hadronization, and develops a mixed shape-multiplicity driven structure in the full-generator configuration. Transporting the individual shower-level functional components downstream shows that shower-level multiplicity information can retain its reweighting power, whereas the corresponding shape responses need not do so even though shape becomes important again at later stages. The jet-mass factors, meanwhile, are limited by poor statistical support. This KAN-based framework therefore provides a functional anatomy of generator-model dependence, exposing both persistent structures and support failures that are hidden inside a single global classifier-derived reweighting function.

Figures & tables

Figure 1: Out-of-fold Pythia - Herwig separation on the common eight-observable representation. The dotted reference corresponds to chance-level separation.

Generator configuration	additive KAN	deep KAN	$\Delta\mathrm{AUC}$
A: shower only	0.9787	0.9924	+0.0137
B: shower + hadronization	0.5933	0.5941	+0.0008
C: full generator	0.6061	0.6062	+0.0001

Table 1: OOF generator separation on the development sample. The final column diagnoses additional non-additive discriminating information.

Figure 2: Representative additive-KAN response functions at stage A.

Figure 3: Representative additive-KAN response functions at stage B.

Figure 4: Representative additive-KAN response functions at stage C.

Panel of Figure 5 — Figure 5: Grouped Shapley allocations for the multiplicity, mass, and shape sectors at the three generator stages. Error bars show hard-event-clustered bootstrap intervals.

Panel of Figure 6 — Figure 6: Transport of the individual stage A factors to the later generator stages. The dashed line shows the unweighted generator separation and the dotted line corresponds to an AUC of $0.5$ . The mass factors are omitted because their effective sample sizes are too small for a reliable weighted comparison.

Figure 7: Development and final-test results for the transported stage A leading-multiplicity factor.

Panel of Figure 8 — Figure 8: Statistical support of the stage A component weights. Left: number of development events required to accumulate $50\%$ of the normalized component weight. Right: leading-jet mass associated with the same influential hard event in the paired Pythia and Herwig configurations at stages A, B, and C.

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Stage	value function	multiplicity	mass	shape
A	separation	$0.4050\,[0.4021,0.4078]$	$-0.0005\,[-0.0034,0.0022]$	$0.0751\,[0.0720,0.0789]$
A	closure	$0.1796\,[0.1403,0.2186]$	$0.0286\,[0.0072,0.0538]$	$-0.0001\,[-0.0048,0.0044]$
B	separation	$0.0039\,[0.0024,0.0056]$	$0.0498\,[0.0442,0.0554]$	$0.0478\,[0.0415,0.0544]$
B	closure	$-0.0024\,[-0.0031,-0.0013]$	$0.0537\,[0.0514,0.0553]$	$0.0378\,[0.0341,0.0392]$
C	separation	$0.0365\,[0.0308,0.0420]$	$0.0205\,[0.0148,0.0257]$	$0.0554\,[0.0488,0.0625]$
C	closure	$0.0352\,[0.0304,0.0390]$	$0.0133\,[0.0104,0.0153]$	$0.0519\,[0.0480,0.0561]$

Appendix

Table 2: Grouped Shapley allocations used in Analysis I. Brackets denote

95\%

hard-event-clustered bootstrap intervals.

Explore similar work

Jan 16, 2026

AllShowers: One model for all calorimeter showers

Accurate and efficient detector simulation is essential for modern collider experiments. To reduce the high computational cost, various fast machine learning surrogate models have been proposed. Traditional surrogate models for calorimeter shower modeling train separate networks for each particle species, limiting scalability and reuse. We introduce AllShowers, a unified generative model that simulates calorimeter showers across multiple particle types using a single generative model. AllShowers is a continuous normalizing flow model with a Transformer architecture, enabling it to generate complex spatial and energy correlations in variable-length point cloud representations of showers. Trained on a diverse dataset of simulated showers in the highly granular ILD detector, the model demonstrates the ability to generate realistic showers for electrons, photons, and charged and neutral hadrons across a wide range of incident energies and angles without retraining. In addition to unifying shower generation for multiple particle types, AllShowers surpasses the fidelity of previous single-particle-type models for hadronic showers. Key innovations include the use of a layer embedding, allowing the model to learn all relevant calorimeter layer properties; a custom attention masking scheme to reduce computational demands and introduce a helpful inductive bias; and a shower- and layer-wise optimal transport mapping to improve training convergence and sample quality. AllShowers marks a significant step towards a universal model for calorimeter shower simulations in collider experiments.

Thorsten Buss, Henry Day-Hall, Frank Gaede +2

Institut für Experimentalphysik, Universität Hamburg, Hamburg, Germany · Deutsches Elektronen-Synchrotron DESY, Hamburg, Germany

Neural Surrogate Modeling Generative Modeling Point Cloud Generation Physics-Based Simulation

Ask PDF

Apr 2, 2026

Generative models on phase space

Deep generative models such as diffusion and flow matching are powerful machine learning tools capable of learning and sampling from high-dimensional distributions. They are particularly useful when the training data appears to be concentrated on a submanifold of the data embedding space. For high-energy physics data, consisting of collections of relativistic energy-momentum 4-vectors, this submanifold can enforce extremely strong physically-motivated priors, such as energy and momentum conservation. If these constraints are learned only approximately, rather than exactly, this can inhibit the interpretability and reliability of such generative models. To remedy this deficiency, we introduce generative models which are, by construction, confined at every step of their sampling trajectory to the manifold of massless N-particle Lorentz-invariant phase space in the center-of-momentum frame. In the case of diffusion models, the "pure noise" forward process endpoint corresponds to the uniform distribution on phase space, which provides a clear starting point from which to identify how correlations among the particles emerge during the reverse (de-noising) process. We demonstrate that our models are able to learn both few-particle and many-particle distributions with various singularity structures, paving the way for future interpretability studies using generative models trained on simulated jet data.

Zachary Bogorad, Ibrahim Elsharkawy, Yonatan Kahn +2

Theory Division, Fermi National Accelerator Laboratory, Batavia, IL, USA · Department of Physics, University of Toronto and Vector Institute, Toronto, ON, Canada and NERSC, Lawrence Berkeley National Laboratory, Berkeley, California, USA · Department of Physics, University of Toronto and Vector Institute, Toronto, ON, Canada +2

Generative Modeling Constrained Generative Modeling Physics-Informed Generative Modeling Diffusion Models High-Energy Physics

Ask PDF

May 27, 2026

Neural Scaling Laws for Jet Generation

Recently observed empirical scaling laws describe the performance of foundation-type models as three independent key quantities -- dataset size, compute, and model parameters -- are modified. Extracting these scaling laws informs the training of large complex models for which the tuning of hyperparameters in traditional ways is not feasible. This work for the first time explores if scaling laws can also be observed for the task of particle jet generation -- both relevant as a pre-training objective for foundation models and as in-situ simulation by itself. We indeed replicate the key logarithmic scaling law behavior for model-size scaling. Beyond studying the next token prediction validation loss of the generative model, we also study the sliced Wasserstein distance of five physical quantities that are not immediately available to the model during training. Our study shows that this quantity is monotonically related to the next token prediction validation loss, meaning that this loss is indeed a good proxy for the physics performance. For the scaling with dataset size and compute, we observe substantially weaker scaling behavior of both the loss and the sliced Wasserstein distance. We analyze this behavior by introducing the concept of a learnable window, and argue that autoregressive next token prediction on jet constituents exhibits comparatively rapid saturation relative to language-model studies. We discuss possible origins of this behavior, including the stochastic nature of QCD radiation and differences between generative and supervised learning tasks in collider physics.

Oz Amram, Darius A. Faroughy, Tjarko Gerdes +5

Fermi National Accelerator Laboratory, Batavia, IL 60510, USA · NHETC, Dept. of Physics and Astronomy, Rutgers University, Piscataway, NJ 08854, USA · Institute for Experimental Physics, Universität Hamburg, Luruper Chaussee 149, 22761 Hamburg, Germany +1

High-Energy Physics Autoregressive Generation Neural Scaling Laws

Ask PDF