arXiv subjects

Andrew J. Larkoski

Publications and source records attributed to Andrew J. Larkoski.

At least 19 recordsLinked to original sources

arXiv2026

More QCD Masterclass Lectures on Jet Physics and Machine Learning

These lectures were presented at the 2026 QCD Masterclass in Saint-Jacut-de-la-mer, France. They supplement the published notes from the 2024 school with presentation of a few new topics that have become popular in recent years. These include event metrics, anomaly detection, scaling laws, the limits of classification performance, and regression that respects spacetime symmetries. Many of the presented results are known, but some new results are derived, especially related to anomaly detection. As was the case in those original notes, while all these problems are motivated from machine learning applied to QCD, no discussion of details of machine learning is presented in these notes. All results follow from application of fundamental principles of QCD. I begin these notes with an apology for humanist physics.

hep-ph↗

arXiv2026

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.

hep-ph↗

arXiv2026

Flavor-Changing Non-Global Logarithms

Non-global logarithms are low energy correlations between the substructure of a jet and the event in which it is immersed. We study the leading non-global logarithms that arise from soft quark--anti-quark emission and calculate their coefficient as a series in the jet radius, $R$, in arbitrary processes. We calculate the exact coefficient through quadratic order in $R$, and show that this truncation is within $5\%$ of the complete result for radii up to $R = 1$. These quark flavor-dependent non-global logarithms are also responsible for the infrared unsafety of a naïve definition of jet flavor that is simply the net sum of quark flavors in the jet of interest. We propose a small modification of this naïve jet flavor that we call subtractive jet flavor in which the problematic soft logarithms are explicitly subtracted. We further demonstrate how our analytic results can be interfaced with automated numerical fixed-order codes to extract subtractive jet flavor cross sections.

hep-ph↗

arXiv2025

Factorization for Collider Dataspace Correlators

A metric on the space of collider physics data enables analysis of its geometrical properties, like dimensionality or curvature, as well as quantifying the density with which a finite, discrete ensemble of data samples the space. We provide the first systematically-improvable precision calculations on this dataspace, presenting predictions resummed to next-to-leading logarithmic accuracy, using the Spectral Energy Mover's Distance (SEMD) as its metric. This is accomplished by demonstration of factorization of soft and collinear contributions to the metric at leading power and renormalization group evolution of the single-scale functions that are present in the factorization theorem. As applications of this general framework, we calculate the two-point correlator between pairs of jets on the dataspace, and the measure of the non-Gaussian fluctuations in a finite dataset. For the non-Gaussianities, our calculations validate the existence of a universal structure that had been previously observed in simulated data. As byproducts of this analysis, we also calculate the two-loop anomalous dimension of the SEMD metric and show that the original Energy Mover's Distance metric is identical to the SEMD through next-to-next-to-leading logarithmic accuracy.

hep-ph↗

arXiv2025

A Step Toward Interpretability: Smearing the Likelihood

The problem of interpretability of machine learning architecture in particle physics has no agreed-upon definition, much less any proposed solution. We present a first modest step toward these goals by proposing a definition and corresponding practical method for isolation and identification of relevant physical energy scales exploited by the machine. This is accomplished by smearing or averaging over all input events that lie within a prescribed metric energy distance of one another and correspondingly renders any quantity measured on a finite, discrete dataset continuous over the dataspace. Within this approach, we are able to explicitly demonstrate that (approximate) scaling laws are a consequence of extreme value theory applied to analysis of the distribution of the irreducible minimal distance over which a machine must extrapolate given a finite dataset. As an example, we study quark versus gluon jet identification, construct the smeared likelihood, and show that discrimination power steadily increases as resolution decreases, indicating that the true likelihood for the problem is sensitive to emissions at all scales.

hep-ph↗

arXiv2025

Non-Gaussianities in Collider Metric Binning

Metrics for rigorously defining a distance between two events have been used to study the properties of the dataspace manifold of particle collider physics. The probability distribution of pairwise distances on this dataspace is unique with probability 1, and so this suggests a method to search for and identify new physics by the deviation of measurement from a null hypothesis prediction. To quantify the deviation statistically, we directly calculate the probability distribution of the number of event pairs that land in the bin a fixed distance apart. This distribution is not generically Gaussian and the ratio of the standard deviation to the mean entries in a bin scales inversely with the square-root of the number of events in the data ensemble. If the dataspace manifold exhibits some enhanced symmetry, the number of entries is Gaussian, and further fluctuations about the mean scale away like the inverse of the number of events. We define a robust measure of the non-Gaussianity of the bin-by-bin statistics of the distance distribution, and demonstrate in simulated data of jets from quantum chromodynamics sensitivity to the parton-to-hadron transition and that the manifold of events enjoys enhanced symmetries as their energy increases.

hep-ph↗

arXiv2025

SPECTER: Efficient Evaluation of the Spectral EMD

The Energy Mover's Distance (EMD) has seen use in collider physics as a metric between events and as a geometric method of defining infrared and collinear safe observables. Recently, the Spectral Energy Mover's Distance (SEMD) has been proposed as a more analytically tractable alternative to the EMD. In this work, we obtain a closed-form expression for the Riemannian-like p = 2 SEMD metric between events, eliminating the need to numerically solve an optimal transport problem. Additionally, we show how the SEMD can be used to define event and jet shape observables by minimizing the distance between events and parameterized energy flows (similar to the EMD), and we obtain closed-form expressions for several of these observables. We also present the SPECTER framework, an efficient and highly parallelized implementation of the SEMD metric and SEMD-derived shape observables as an analogue of the previously-introduced SHAPER for EMD-based computations. We demonstrate that computing the SEMD with SPECTER can be up to a thousand times faster than computing the EMD with standard optimal transport libraries.

hep-ph↗

arXiv2024

Systematically Constructing the Likelihood for Boosted $H\to gg$ Decays

We study the binary discrimination problem of identification of boosted $H\to gg$ decays from massive QCD jets in a systematic expansion in the strong coupling. Though this decay mode of the Higgs is unlikely to be discovered at the LHC, we analytically demonstrate several features of the likelihood ratio for this problem through explicit analysis of signal and background matrix elements. Through leading-order, we prove that by imposing a constraint on the jet mass and measuring the energy fraction of the softer subjet an improvement of signal to background ratio that is independent of the kinematics of the jets at high boosts can be obtained, and is approximately equal to the inverse of the strong coupling evaluated at the Higgs mass. At next-to-leading order, we construct a powerful discrimination observable through a sort of anomaly detection approach by simply inverting the next-to-leading order $H\to gg$ matrix element with soft gluon emission, which is naturally infrared and collinear safe. Our analytic conclusions are validated in simulated data from all-purpose event generators and subsequent parton showering and demonstrate that the signal-to-background ratio can be improved by a factor of several hundred at high, but accessible, jet energies at the LHC.

hep-ph↗

arXiv2024

Systematic Interpretability and the Likelihood for Boosted Top Quark Identification

Identification of boosted, hadronically-decaying top quarks is a problem of central importance for physics goals of the Large Hadron Collider. We present a theoretical analysis of top quark tagging, establishing zeroth-order, minimal assumptions that should be satisfied by any purported top-tagged jet, like existence of three hard subjets, a bottom-tagged subjet, total mass consistent with the top quark, and a pairwise subjet mass consistent with the W boson. From these minimal assumptions, we construct the optimal discrimination observable, the likelihood ratio, for the binary discrimination problem of top quark-initiated versus bottom quark-initiated jets through next-to-leading order in the strong coupling. We compare and compute corresponding signal and background efficiencies both analytically and from simulated data, validating an understanding of the relevant physics identified and exploited by the likelihood. In the process, we construct a method for systematic interpretability of the likelihood ratio for this problem, and explicitly establish a hard floor on possible discrimination power. These results can correspondingly be applied to understanding and interpreting machine learning studies of this problem.

hep-ph↗

arXiv2024

Flavor Factorization at Two-Loops

Recently, a factorization theorem was proposed for partonic flavor evolution as defined by the net flavor of the Winner-Take-All axis of a jet. We validate the factorization theorem through explicit calculation at two-loop order, and in the process extract all anomalous dimensions and renormalization factors for any ultraviolet-to-infrared flavor transition at this order. These results can then be used to extract the renormalized hard function for flavored jet production at next-to-next-to-leading order for any process of interest.

hep-ph↗

arXiv2024

QCD Masterclass Lectures on Jet Physics and Machine Learning

These lectures were presented at the 2024 QCD Masterclass in Saint-Jacut-de-la-Mer, France. They introduce and review fundamental theorems and principles of machine learning within the context of collider particle physics, focused on application to jet identification and discrimination. Numerous examples of binary discrimination in jet physics are studied in detail, including $H\to b\bar b$ identification in fixed-order perturbation theory, generic one-versus two-prong discrimination with parametric power counting techniques, and up versus down quark jet classification by assuming the central limit theorem, isospin conservation, and a convergent moment expansion of the single particle energy distribution. Quark versus gluon jet discrimination is considered in multiple contexts, from using additive, infrared and collinear safe observables, to using hadronic multiplicity, and to including measurements of the jet charge. While many of the results presented here are well known, some novel results are presented, the most prominent being a parametrized expression for the likelihood ratio of quark versus gluon discrimination for jets on which hadronic multiplicity and jet charge are simultaneously measured. End-of-lecture exercises are also provided.

hep-ph↗

arXiv2024

The Phase Space Distance Between Collider Events

How can one fully harness the power of physics encoded in relativistic $N$-body phase space? Topologically, phase space is isomorphic to the product space of a simplex and a hypersphere and can be equipped with explicit coordinates and a Riemannian metric. This natural structure that scaffolds the space on which all collider physics events live opens up new directions for machine learning applications and implementation. Here we present a detailed construction of the phase space manifold and its differential line element, identifying particle ordering prescriptions that ensure that the metric satisfies necessary properties. We apply the phase space metric to several binary classification tasks, including discrimination of high-multiplicity resonance decays or boosted hadronic decays of electroweak bosons from QCD processes, and demonstrate powerful performance on simulated data. Our work demonstrates the many benefits of promoting phase space from merely a background on which calculations take place to being geometrically entwined with a theory's dynamics.

hep-ph↗

arXiv2024

Is infrared-collinear safe information all you need for jet classification?

Machine learning-based jet classifiers are able to achieve impressive tagging performance in a variety of applications in high-energy and nuclear physics. However, it remains unclear in many cases which aspects of jets give rise to this discriminating power, and whether jet observables that are tractable in perturbative QCD such as those obeying infrared-collinear (IRC) safety serve as sufficient inputs. In this article, we introduce a new classifier, Jet Flow Networks (JFNs), in an effort to address the question of whether IRC unsafe information provides additional discriminating power in jet classification. JFNs are permutation-invariant neural networks (deep sets) that take as input the kinematic information of reconstructed subjets. The subjet radius and a cut on the subjet's transverse momenta serve as tunable hyperparameters enabling a controllable sensitivity to soft emissions and nonperturbative effects. We demonstrate the performance of JFNs for quark vs. gluon and Z vs. QCD jet tagging. For small subjet radii and transverse momentum cuts, the performance of JFNs is equivalent to the IRC-unsafe Particle Flow Networks (PFNs), demonstrating that infrared-collinear unsafe information is not necessary to achieve strong discrimination for both cases. As the subjet radius is increased, the performance of the JFNs remains essentially unchanged until physical thresholds that we identify are crossed. For relatively large subjet radii, we show that the JFNs may offer an increased model independence with a modest tradeoff in performance compared to classifiers that use the full particle information of the jet. These results shed new light on how machines learn patterns in high-energy physics data

hep-ph↗

arXiv2024

Analytical Insights on Hadronic Top Quark Polarimetry

Top quark polarization provides an important tool for studying its production mechanisms, spin correlations, top quark properties, and new physics searches. Unlike lighter quarks, the top quark's polarization remains intact until its decay, enabling precise spin measurements. While the down-type fermions from $W$ boson decay are known to be effective spin analyzers, charged leptons have typically been the main target for most analyses. In this paper, we investigate the relevance of global jet dynamics -- considering kinematics, jet charges, and particle multiplicity -- for hadronic top quark polarimetry. The formalism used allows for analytical derivations obtained throughout the manuscript, offering deeper insights into the corresponding phenomenology.

hep-ph↗

arXiv2024

Hadronic Top Quark Polarimetry with ParticleNet

Precision studies for top quark physics are a cornerstone of the Large Hadron Collider program. Polarization, probed through decay kinematics, provides a unique tool to scrutinize the top quark across its various production modes and to explore potential new physics effects. However, the top quark most often decays hadronically, for which unambiguous identification of its decay products sensitive to top quark polarization is not possible. In this Letter, we introduce a jet flavor tagging method to significantly improve spin analyzing power in hadronic decays, going beyond exclusive kinematic information employed in previous studies. We provide parametric estimates of the improvement from flavor tagging with any set of measured observables and demonstrate this in practice on simulated data using a Graph Neural Network (GNN). We find that the spin analyzing power in hadronic decays can improve by approximately 20% (40%) compared to the kinematic approach, assuming an efficiency of 0.5 (0.2) for the network.

hep-ph↗

arXiv2024

Binary Discrimination Through Next-to-Leading Order

Binary discrimination between well-defined signal and background datasets is a problem of fundamental importance in particle physics. With detailed event simulation and the advent of extensive deep learning tools, identification of the likelihood ratio has typically been reserved as a computational problem. However, this approach can obscure overtraining or excessive sensitivity to tuned features of the simulation that may not be well-defined theoretically. Here, we present the first analysis of binary discrimination for signal and background distributions for which their likelihood ratio is infrared and collinear safe, and can therefore be calculated order-by-order in perturbation theory. We present explicit, general formulas for receiver operator characteristic curves and the area under it through next-to-leading order. As a demonstration of this formalism, we apply it to discrimination of highly-boosted Higgs decays from gluon splitting to bottom quarks. Effects at next-to-leading order are first sensitive to the flow of color in the jet and significantly modify discrimination performance at leading-order. In the limit of infinite boost, these events can be perfectly discriminated because only the gluon will radiate at finite angles from the bottom quarks, and we find that large effects persist at energies accessible at the Large Hadron Collider. Next-to-leading order is therefore required to qualitatively understand results using machine-learning methods.

hep-ph↗

arXiv2024

Flavor Fragmentation Function Factorization

A definition of partonic jet flavor that is both theoretically well-defined and experimentally robust would have profound implications for measurements and predictions especially for heavy flavor applications. Recently, a definition of jet flavor was introduced as the net flavor flowing along the direction of the Winner-Take-All axis of a jet which is soft safe to all orders, but not collinear safe. Here, we exploit the lack of collinear safety and propose a factorization theorem of perturbative flavor fragmentation functions that resum collinear divergences and describe the evolution of flavor from the short distance of jet production to the long distance at which hadronization occurs. Collinear flavor evolution is governed by a small modification of the DGLAP equations. We present a detailed all-orders analysis and identify exact relations that must hold amongst the various anomalous dimensions by probability conservation and the existence of fixed points of the renormalization group flow. We explicitly validate the factorization theorem at one-loop order, and demonstrate its consistency at two loops in particular flavor channels. Starting at two-loops, constraints on phase space imposed by flavor measurements potentially allow for non-trivial soft contributions, but we demonstrate that they are scaleless and so explicitly vanish, ensuring that soft particles are summed inclusively and all divergences are exclusively collinear in nature. This factorization theorem opens the door to precision calculations with identified flavor in the infrared.

hep-ph↗

arXiv2023

A Spectral Metric for Collider Geometry

By quantifying the distance between two collider events, one can triangulate a metric space and reframe collider data analysis as computational geometry. One popular geometric approach is to first represent events as an energy flow on an idealized celestial sphere and then define the metric in terms of optimal transport in two dimensions. In this paper, we advocate for representing events in terms of a spectral function that encodes pairwise particle angles and products of particle energies, which enables a metric distance defined in terms of one-dimensional optimal transport. This approach has the advantage of automatically incorporating obvious isometries of the data, like rotations about the colliding beam axis. It also facilitates first-principles calculations, since there are simple closed-form expressions for optimal transport in one dimension. Up to isometries and event sets of measure zero, the spectral representation is unique, so the metric on the space of spectral functions is a metric on the space of events. At lowest order in perturbation theory in electron-positron collisions, our metric is simply the summed squared invariant masses of the two event hemispheres. Going to higher orders, we present predictions for the distribution of metric distances between jets in fixed-order and resummed perturbation theory as well as in parton-shower generators. Finally, we speculate on whether the spectral approach could furnish a useful metric on the space of quantum field theories.

hep-ph↗