arXiv subjects

Michael Krämer

Publications and source records attributed to Michael Krämer.

At least 37 records · Page 2Linked to original sources

arXiv2024

NNLL-fast 2.0: Coloured Sparticle Production at the LHC Run 3 with $\sqrt{S}$ = 13.6 TeV

We report on updated precision predictions for total cross sections of coloured supersymmetric particle production at the LHC with a centre-of-mass energy of $\sqrt S$ = 13.6 TeV, computed with the modern PDF4LHC21 set. The cross sections are calculated at an approximated NNLO accuracy in QCD and contain corrections from the threshold resummation of soft-gluon emission up to NNLL accuracy as well as Coulomb-gluon contributions including bound-state terms. The corrections are found to increase the cross sections and reduce the theoretical uncertainty as compared to the best available fixed-order calculations. These predictions constitute the state-of-the-art calculations and update the existing results for $\sqrt S$ = 13 TeV. We make our new results publicly available in the version 2.0 update to the code package NNLL-fast.

hep-ph↗

arXiv2024

Accurate and robust methods for direct background estimation in resonant anomaly detection

Resonant anomaly detection methods have great potential for enhancing the sensitivity of traditional bump hunt searches. A key component of these methods is a high quality background template used to produce an anomaly score. Using the LHC Olympics R&D dataset, we demonstrate that this background template can also be repurposed to directly estimate the background expectation in a simple cut and count setup. In contrast to a traditional bump hunt, no fit to the invariant mass distribution is needed, thereby avoiding the potential problem of background sculpting. Furthermore, direct background estimation allows working with large background rejection rates, where resonant anomaly detection methods typically show their greatest improvement in significance.

hep-ph↗

arXiv2024

Flavoured Majorana Dark Matter then and now: From freeze-out scenarios to LHC signatures

We study a simplified Dark Matter model in the Dark Minimal Flavour Violation framework. Our model complements the Standard Model with a flavoured Dark Matter Majorana triplet and a coloured scalar mediator that share a Yukawa coupling with the right-handed up-type quarks with the coupling matrix $λ$. We extend previous work on this topic by exploring a large range of cosmologically viable parameter space, including the coannihilation region and, in particular, the region of conversion-driven freeze-out, while considering constraints from $D^0-\bar D^0$ mixing as well as constraints from direct and indirect Dark Matter searches. We find various realisations of conversion-driven freeze-out within the model, that open up allowed windows of parameter space towards small mass splittings and very weak Dark Matter couplings. Finally, we probe the model by reinterpreting current LHC searches for missing energy and long-lived particles. We point out gaps in the coverage of current constraints as well as new opportunities to search for the model at the LHC, in particular, the charge asymmetry in single-top production associated with jets and missing energy.

hep-ph↗

arXiv2024

Cluster Scanning: a novel approach to resonance searches

We propose a new model-independent method for new physics searches called Cluster Scanning. It uses the k-means algorithm to perform clustering in the space of low-level event or jet observables, and separates potentially anomalous clusters to construct a signal-enriched region. The spectra of a selected observable (e.g. invariant mass) in these two regions are then used to determine whether a resonant signal is present. A pseudo-analysis on the LHC Olympics dataset with a $Z'$ resonance shows that Cluster Scanning outperforms the widely used 4-parameter functional background fitting procedures, reducing the number of signal events needed to reach a $3σ$ significant access by a factor of 0.61. Emphasis is placed on the speed of the method, which allows the test statistic to be calibrated on synthetic data.

hep-ph↗

arXiv2024

Critical feature learning in deep neural networks

A key property of neural networks driving their success is their ability to learn features from data. Understanding feature learning from a theoretical viewpoint is an emerging field with many open questions. In this work we capture finite-width effects with a systematic theory of network kernels in deep non-linear neural networks. We show that the Bayesian prior of the network can be written in closed form as a superposition of Gaussian processes, whose kernels are distributed with a variance that depends inversely on the network width N . A large deviation approach, which is exact in the proportional limit for the number of data points $P = αN \rightarrow \infty$, yields a pair of forward-backward equations for the maximum a posteriori kernels in all layers at once. We study their solutions perturbatively to demonstrate how the backward propagation across layers aligns kernels with the target. An alternative field-theoretic formulation shows that kernel adaptation of the Bayesian posterior at finite-width results from fluctuations in the prior: larger fluctuations correspond to a more flexible network prior and thus enable stronger adaptation to data. We thus find a bridge between the classical edge-of-chaos NNGP theory and feature learning, exposing an intricate interplay between criticality, response functions, and feature scale.

cond-mat.dis-nn↗

arXiv2023

Comprehensive exploration of t-channel simplified models of dark matter

We analyse six classes of t-channel dark matter simplified models in which the Standard Model field content is extended by a coloured mediator and a dark matter state. The two new states are enforced to be odd under a new parity, while all Standard Model fields are taken even so that dark matter stability is guaranteed. We study several possibilities for the spin of the new particles and the self-conjugate property of the dark matter, and we focus on model configurations in which the dark matter couples to the right-handed up quark for simplicity. We investigate how the parameter spaces of the six models can be constrained by current and future cosmological, astrophysical and collider searches, and we highlight the strong complementary between those probes. Our results demonstrate that scenarios featuring a complex (non self-conjugate) dark matter field are excluded by cosmology and astrophysics alone, the only possibility to avoid these bounds being to invoke very weak couplings and mechanisms such as conversion-driven freeze-out. For models with self-conjugate dark matter, mediator and dark matter masses are pushed deep into the TeV regime, with the lower limits on the mediator mass reaching 3 to 4 TeV and those on the dark matter mass 1 to 2 TeV. In large parts of the parameter space these strong bounds are driven by same-sign mediator pair production, a channel so far not considered in the experimental analyses embedding t-channel dark matter model interpretations.

hep-ph↗

arXiv2023

A theory of data variability in Neural Network Bayesian inference

Bayesian inference and kernel methods are well established in machine learning. The neural network Gaussian process in particular provides a concept to investigate neural networks in the limit of infinitely wide hidden layers by using kernel and inference methods. Here we build upon this limit and provide a field-theoretic formalism which covers the generalization properties of infinitely wide networks. We systematically compute generalization properties of linear, non-linear, and deep non-linear networks for kernel matrices with heterogeneous entries. In contrast to currently employed spectral methods we derive the generalization properties from the statistical properties of the input, elucidating the interplay of input dimensionality, size of the training data set, and variability of the data. We show that data variability leads to a non-Gaussian action reminiscent of a ($φ^3+φ^4$)-theory. Using our formalism on a synthetic task and on MNIST we obtain a homogeneous kernel matrix approximation for the learning curve as well as corrections due to data variability which allow the estimation of the generalization properties and exact results for the bounds of the learning curves in the case of infinitely many training data points.

cond-mat.dis-nn↗

arXiv2023

What's Anomalous in LHC Jets?

Searches for anomalies are a significant motivation for the LHC and help define key analysis steps, including triggers. We discuss specific examples how LHC anomalies can be defined through probability density estimates, evaluated in a physics space or in an appropriate neural network latent space, and discuss the model-dependence in choosing an appropriate data parameterisation. We illustrate this for classical k-means clustering, a Dirichlet variational autoencoder, and invertible neural networks. For two especially challenging scenarios of jets from a dark sector we evaluate the strengths and limitations of each method.

hep-ph↗

arXiv2023

Back To The Roots: Tree-Based Algorithms for Weakly Supervised Anomaly Detection

Weakly supervised methods have emerged as a powerful tool for model-agnostic anomaly detection at the Large Hadron Collider (LHC). While these methods have shown remarkable performance on specific signatures such as di-jet resonances, their application in a more model-agnostic manner requires dealing with a larger number of potentially noisy input features. In this paper, we show that using boosted decision trees as classifiers in weakly supervised anomaly detection gives superior performance compared to deep neural networks. Boosted decision trees are well known for their effectiveness in tabular data analysis. Our results show that they not only offer significantly faster training and evaluation times, but they are also robust to a large number of noisy input features. By using advanced gradient boosted decision trees in combination with ensembling techniques and an extended set of features, we significantly improve the performance of weakly supervised methods for anomaly detection at the LHC. This advance is a crucial step towards a more model-agnostic search for new physics.

hep-ph↗

arXiv2023

Learning the language of QCD jets with transformers

Transformers have become the primary architecture for natural language processing. In this study, we explore their use for auto-regressive density estimation in high-energy jet physics, which involves working with a high-dimensional space. We draw an analogy between sentences and words in natural language and jets and their constituents in high-energy physics. Specifically, we investigate density estimation for light QCD jets and hadronically decaying boosted top jets. Since transformers allow easy sampling from learned densities, we exploit their generative capability to assess the quality of the density estimate. Our results indicate that the generated data samples closely resemble the original data, as evidenced by the excellent agreement of distributions such as particle multiplicity or jet mass. Furthermore, the generated samples are difficult to distinguish from the original data, even by a powerful supervised classifier. Given their exceptional data processing capabilities, transformers could potentially be trained directly on the massive LHC data sets to learn the probability densities in high-energy jet physics.

hep-ph↗

arXiv2023

A Normalized Autoencoder for LHC Triggers

Autoencoders are an effective analysis tool for the LHC, as they represent one of its main goal of finding physics beyond the Standard Model. The key challenge is that out-of-distribution anomaly searches based on the compressibility of features do not apply to the LHC, while existing density-based searches lack performance. We present the first autoencoder which identifies anomalous jets symmetrically in the directions of higher and lower complexity. The normalized autoencoder combines a standard bottleneck architecture with a well-defined probabilistic description. It works better than all available autoencoders for top vs QCD jets and reliably identifies different dark-jet signals.

hep-ph↗

arXiv2023

Searching for dark matter subhalos in the Fermi-LAT catalog with Bayesian neural networks

About a third of the $γ$-ray sources detected by the Fermi Large Area Telescope (Fermi-LAT) remain unidentified, and some of these could be exotic objects such as dark matter subhalos. We present a search for these sources using Bayesian neural network classification methods applied to the latest 4FGL-DR3 Fermi-LAT catalog. We first simulate the $γ$-ray properties of dark matter subhalos using models from N-body simulations and semi-analytical approaches to the subhalo distribution. We then assess the detectability of this sample in the 4FGL-DR3 catalog using the Fermi-LAT analysis tools. We train our Bayesian neural network to identify candidate dark matter subhalos among the unidentified sources in the 4FGL-DR3 catalog. Our results allow us to derive conservative bounds on the dark matter annihilation cross section by excluding unidentified sources classified as astrophysical-like by our networks. We estimate the number of candidate dark matter subhalos for different dark matter masses and provide a publicly available list for further investigation. Our bounds on the dark matter annihilation cross section are comparable to previous results and become particularly competitive at high dark matter masses.

astro-ph.HE↗

arXiv2022

Unified field theoretical approach to deep and recurrent neuronal networks

Understanding capabilities and limitations of different network architectures is of fundamental importance to machine learning. Bayesian inference on Gaussian processes has proven to be a viable approach for studying recurrent and deep networks in the limit of infinite layer width, $n\to\infty$. Here we present a unified and systematic derivation of the mean-field theory for both architectures that starts from first principles by employing established methods from statistical physics of disordered systems. The theory elucidates that while the mean-field equations are different with regard to their temporal structure, they yet yield identical Gaussian kernels when readouts are taken at a single time point or layer, respectively. Bayesian inference applied to classification then predicts identical performance and capabilities for the two architectures. Numerically, we find that convergence towards the mean-field theory is typically slower for recurrent networks than for deep networks and the convergence speed depends non-trivially on the parameters of the weight prior as well as the depth or number of time steps, respectively. Our method exposes that Gaussian processes are but the lowest order of a systematic expansion in $1/n$ and we compute next-to-leading-order corrections which turn out to be architecture-specific. The formalism thus paves the way to investigate the fundamental differences between recurrent and deep architectures at finite widths $n$.

cond-mat.dis-nn↗

arXiv2022

Classification of Fermi-LAT blazars with Bayesian neural networks

The use of Bayesian neural networks is a novel approach for the classification of gamma-ray sources. We focus on the classification of Fermi-LAT blazar candidates, which can be divided into BL Lacertae objects and Flat Spectrum Radio Quasars. In contrast to conventional dense networks, Bayesian neural networks provide a reliable estimate of the uncertainty of the network predictions. We explore the correspondence between conventional and Bayesian neural networks and the effect of data augmentation. We find that Bayesian neural networks provide a robust classifier with reliable uncertainty estimates and are particularly well suited for classification problems that are based on comparatively small and imbalanced data sets. The results of our blazar candidate classification are valuable input for population studies aimed at constraining the blazar luminosity function and to guide future observational campaigns.

astro-ph.HE↗

arXiv2022

Boosting mono-jet searches with model-agnostic machine learning

We show how weakly supervised machine learning can improve the sensitivity of LHC mono-jet searches to new physics models with anomalous jet dynamics. The Classification Without Labels (CWoLa) method is used to extract all the information available from low-level detector information without any reference to specific new physics models. For the example of a strongly interacting dark matter model, we employ simulated data to show that the discovery potential of an existing generic search can be boosted considerably.

hep-ph↗

arXiv2022

Simple and statistically sound recommendations for analysing physical theories

Physical theories that depend on many parameters or are tested against data from many different experiments pose unique challenges to statistical inference. Many models in particle physics, astrophysics and cosmology fall into one or both of these categories. These issues are often sidestepped with statistically unsound ad hoc methods, involving intersection of parameter intervals estimated by multiple experiments, and random or grid sampling of model parameters. Whilst these methods are easy to apply, they exhibit pathologies even in low-dimensional parameter spaces, and quickly become problematic to use and interpret in higher dimensions. In this article we give clear guidance for going beyond these procedures, suggesting where possible simple methods for performing statistically sound inference, and recommendations of readily-available software tools and standards that can assist in doing so. Our aim is to provide any physicists lacking comprehensive statistical training with recommendations for reaching correct scientific conclusions, with only a modest increase in analysis burden. Our examples can be reproduced with the code publicly available at https://doi.org/10.5281/zenodo.4322283.

hep-ph↗

arXiv2022

Integrating Dark Matter, Modified Gravity, and the Humanities

Editorial of a special issue on dark matter & modified gravity, distributed across the journals Studies in History and Philosophy of Modern Physics and Studies in History and Philosophy of Science. Published version of the open access editorial (in SHPS) available here: https://doi.org/10.1016/j.shpsa.2021.08.015. The six papers are collected here: https://www.sciencedirect.com/journal/studies-in-history-and-philosophy-of-science-part-b-studies-in-history-and-philosophy-of-modern-physics/special-issue/10CR71RJLWM.

physics.hist-ph↗

arXiv2021

Constraining dark matter annihilation with cosmic ray antiprotons using neural networks

The interpretation of data from indirect detection experiments searching for dark matter annihilations requires computationally expensive simulations of cosmic-ray propagation. In this work we present a new method based on Recurrent Neural Networks that significantly accelerates simulations of secondary and dark matter Galactic cosmic ray antiprotons while achieving excellent accuracy. This approach allows for an efficient profiling or marginalisation over the nuisance parameters of a cosmic ray propagation model in order to perform parameter scans for a wide range of dark matter models. We identify importance sampling as particularly suitable for ensuring that the network is only evaluated in well-trained parameter regions. We present resulting constraints using the most recent AMS-02 antiproton data on several models of Weakly Interacting Massive Particles. The fully trained networks are released as DarkRayNet together with this work and achieve a speed-up of the runtime by at least two orders of magnitude compared to conventional approaches.

astro-ph.HE↗