Nadezda Chernyavskaya

There has been a recent explosion in research into machine-learning-based generative modeling to tackle computational challenges for simulations in high energy physics (HEP). In order to use such alternative simulators in practice, we need well-defined metrics to compare different generative models and evaluate their discrepancy from the true distributions. We present the first systematic review and investigation into evaluation metrics and their sensitivity to failure modes of generative models, using the framework of two-sample goodness-of-fit testing, and their relevance and viability for HEP. Inspired by previous work in both physics and computer vision, we propose two new metrics, the Fr\'echet and kernel physics distances (FPD and KPD, respectively) and perform a variety of experiments measuring their performance on simple Gaussian-distributed and simulated high energy jet datasets. We find FPD, in particular, to be the most sensitive metric to all alternative jet distributions tested and recommend its adoption, along with the KPD and Wasserstein distances between individual feature distributions, for evaluating generative models in HEP. We finally demonstrate the efficacy of these proposed metrics in evaluating and comparing a novel attention-based generative adversarial particle transformer to the state-of-the-art message-passing generative adversarial network jet simulation model. The code for our proposed metrics is provided in the open source jetnet python library.

There has been significant work recently in developing machine learning models in high energy physics (HEP), for tasks such as classification, simulation, and anomaly detection. Typically, these models are adapted from those designed for datasets in computer vision or natural language processing without necessarily incorporating inductive biases suited to HEP data, such as respecting its inherent symmetries. Such inductive biases can make the model more performant and interpretable, and reduce the amount of training data needed. To that end, we develop the Lorentz group autoencoder (LGAE), an autoencoder model equivariant with respect to the proper, orthochronous Lorentz group $\mathrm{SO}^+(3,1)$, with a latent space living in the representations of the group. We present our architecture and several experimental results on jets at the LHC and find it significantly outperforms a non-Lorentz-equivariant graph neural network baseline on compression and reconstruction, and anomaly detection. We also demonstrate the advantage of such an equivariant model in analyzing the latent space of the autoencoder, which can have a significant impact on the explainability of anomalies found by such black-box machine learning models.

There has been significant work recently in developing machine learning (ML) models in high energy physics (HEP) for tasks such as classification, simulation, and anomaly detection. Often these models are adapted from those designed for datasets in computer vision or natural language processing, which lack inductive biases suited to HEP data, such as equivariance to its inherent symmetries. Such biases have been shown to make models more performant and interpretable, and reduce the amount of training data needed. To that end, we develop the Lorentz group autoencoder (LGAE), an autoencoder model equivariant with respect to the proper, orthochronous Lorentz group $\mathrm{SO}^+(3,1)$, with a latent space living in the representations of the group. We present our architecture and several experimental results on jets at the LHC and find it outperforms graph and convolutional neural network baseline models on several compression, reconstruction, and anomaly detection metrics. We also demonstrate the advantage of such an equivariant model in analyzing the latent space of the autoencoder, which can improve the explainability of potential anomalies discovered by such ML models.

We investigate how to improve new physics detection strategies exploiting variational autoencoders and normalizing flows for anomaly detection at the Large Hadron Collider. As a working example, we consider the DarkMachines challenge dataset. We show how different design choices (e.g., event representations, anomaly score definitions, network architectures) affect the result on specific benchmark new physics models. Once a baseline is established, we discuss how to improve the anomaly detection accuracy by exploiting normalizing flow layers in the latent space of the variational autoencoder.

There has been a recent explosion in research into machine-learning-based generative modeling to tackle computational challenges for simulations in high energy physics (HEP). In order to use such alternative simulators in practice, we need well defined metrics to compare different generative models and evaluate their discrepancy from the true distributions. We present the first systematic review and investigation into evaluation metrics and their sensitivity to failure modes of generative models, using the framework of two-sample goodness-of-fit testing, and their relevance and viability for HEP. Inspired by previous work in both physics and computer vision, we propose two new metrics, the Fr\'echet and kernel physics distances (FPD and KPD), and perform a variety of experiments measuring their performance on simple Gaussian-distributed, and simulated high energy jet datasets. We find FPD, in particular, to be the most sensitive metric to all alternative jet distributions tested and recommend its adoption, along with the KPD and Wasserstein distances between individual feature distributions, for evaluating generative models in HEP. We finally demonstrate the efficacy of these proposed metrics in evaluating and comparing a novel attention-based generative adversarial particle transformer to the state-of-the-art message-passing generative adversarial network jet simulation model.

We study how to use deep variational autoencoders (VAEs) for a fast simulation of jets of particles at the Large Hadron Collider. We represent jets as a list of constituents, characterized by their momenta. Starting from a simulation of the jet before detector effects, we train a deep VAE to return the corresponding list of constituents after detection. Doing so, we bypass both the time-consuming detector simulation and the collision reconstruction steps of a traditional processing chain, speeding up significantly the events generation workflow. Through model optimization and hyperparameter tuning, we achieve state-of-the-art precision on the jet four-momentum, while providing an accurate description of the constituents momenta, and an inference time comparable to that of a rule-based fast simulation.

Abstract We present the current stage of research progress towards a one-pass, completely Machine Learning (ML) based imaging calorimeter reconstruction. The model used is based on Graph Neural Networks (GNNs) and directly analyzes the hits in each HGCAL endcap. The ML algorithm is trained to predict clusters of hits originating from the same incident particle by labeling the hits with the same cluster index. We impose simple criteria to assess whether the hits associated as a cluster by the prediction are matched to those hits resulting from any particular individual incident particles. The algorithm is studied by simulating two tau leptons in each of the two HGCAL endcaps, where each tau may decay according to its measured standard model branching probabilities. The simulation includes the material interaction of the tau decay products which may create additional particles incident upon the calorimeter. Using this varied multiparticle environment we can investigate the application of this reconstruction technique and begin to characterize energy containment and performance.

Transverse momentum spectra, $$\mathrm{d}^2\sigma /(\mathrm{d}\eta \mathrm{d}p_T^2)$$ , of charged hadron production in $$pp$$ -collisions are considered in terms of a recently introduced two component model. The shapes of the particle distributions vary as a function of the c.m.s. energy in the collision and the measured pseudorapidity interval. As a result the pseudorapidity of a secondary hadron in the moving proton rest frame is shown to be a universal parameter describing the shape of the spectra in pp-collisions. In order to extract predictions on the double-differential cross sections $$\mathrm{d}^2\sigma /(\mathrm{d}\eta \mathrm{d}p_T^2)$$ of hadron production for future LHC-measurements the different sets of available experimental data have been used in this study.

Abstract We present an end-to-end reconstruction algorithm to build particle candidates from detector hits in next-generation granular calorimeters similar to that foreseen for the high-luminosity upgrade of the CMS detector. The algorithm exploits a distance-weighted graph neural network, trained with object condensation, a graph segmentation technique. Through a single-shot approach, the reconstruction task is paired with energy regression. We describe the reconstruction performance in terms of efficiency as well as in terms of energy resolution. In addition, we show the jet reconstruction performance of our method and discuss its inference computational cost. To our knowledge, this work is the first-ever example of single-shot calorimetric reconstruction of $${\mathcal {O}}(1000)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>O</mml:mi> <mml:mo>(</mml:mo> <mml:mn>1000</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:math> particles in high-luminosity conditions with 200 pileup.

Abstract Confining dark sectors with pseudo-conformal dynamics can produce Soft Unclustered Energy Patterns (SUEP), at the Large Hadron Collider: the production of dark quarks in proton–proton collisions leading to a dark shower and the high-multiplicity production of dark hadrons. The final experimental signature is spherically symmetric energy deposits by an anomalously large number of soft Standard Model particles with a transverse energy of O(100) $$\,\text {MeV}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mspace /> <mml:mtext>MeV</mml:mtext> </mml:mrow> </mml:math> . Assuming Yukawa-like couplings of the scalar portal state, the dominant production mode is gluon fusion, and the dominant background comes from multi-jet QCD events. We have developed a deep learning-based Anomaly Detection technique to reject QCD jets and identify any anomalous signature, including SUEP, in real-time in the High-Level Trigger system of experiments like the Compact Muon Solenoid at the Large Hadron Collider. A deep convolutional neural autoencoder network has been trained using QCD events by taking transverse energy deposits in the inner tracker, electromagnetic calorimeter, and hadron calorimeter sub-detectors as 3-channel image data. Due to the sparse nature of the data, only $$\sim $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mo>∼</mml:mo> </mml:math> 0.5% of the total $$\sim $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mo>∼</mml:mo> </mml:math> $${300}\,\textrm{k}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mn>300</mml:mn> <mml:mspace /> <mml:mtext>k</mml:mtext> </mml:mrow> </mml:math> image pixels have nonzero values. To tackle this challenge, a nonstandard loss function, the inverse of the so-called Dice Loss, is exploited. The trained autoencoder with learned spatial features of QCD jets can detect 40% of the SUEP events, with a QCD event mistagging rate as low as 2%. The model inference time has been measured using the $$^{\texttt {TM}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mrow /> <mml:mi>TM</mml:mi> </mml:msup> </mml:math> processor and found to be $$\sim $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mo>∼</mml:mo> </mml:math> $${20}\textrm{ms}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mn>20</mml:mn> <mml:mtext>ms</mml:mtext> </mml:mrow> </mml:math> , which perfectly satisfies the High-Level Trigger system’s latency of $$\mathcal {O}(10^2)~\textrm{ms}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>O</mml:mi> <mml:mo>(</mml:mo> <mml:msup> <mml:mn>10</mml:mn> <mml:mn>2</mml:mn> </mml:msup> <mml:mo>)</mml:mo> <mml:mspace /> <mml:mtext>ms</mml:mtext> </mml:mrow> </mml:math> . Given the virtue of the unsupervised learning of the autoencoders, the trained model can be applied to any new physics model that predicts an experimental signature anomalous to QCD jets.

Recently, transformer-based foundation models have proven to be a generalized architecture applicable to various data modalities, ranging from text to audio and even a combination of multiple modalities. Transformers by design should accurately model the non-trivial structure of particle showers thanks to the absence of strong inductive bias, better modeling of long-range dependencies, and interpolation and extrapolation capabilities. In this paper, we explore a transformer-based generative model for detector-agnostic fast shower simulation, where the goal is to generate synthetic particle showers, i.e., the energy depositions in the calorimeter. When trained with an adequate amount and variety of showers, these models should learn better representations compared to other deep learning models, and hence should quickly adapt to new detectors. In this work, we will show the prototype of a transformer-based generative model for fast shower simulation, as well as explore certain aspects of transformer architecture such as input data representation, sequence formation, and the learning mechanism for our unconventional shower data.

The observation of Higgs boson production in association with a top quark-antiquark pair is reported, based on a combined analysis of proton-proton collision data at center-of-mass energies of √s = 7, 8, and 13 TeV, corresponding to integrated luminosities of up to 5.1, 19.7, and 35.9  fb^(-1), respectively. The data were collected with the CMS detector at the CERN LHC. The results of statistically independent searches for Higgs bosons produced in conjunction with a top quark-antiquark pair and decaying to pairs of W bosons, Z bosons, photons, τ leptons, or bottom quark jets are combined to maximize sensitivity. An excess of events is observed, with a significance of 5.2 standard deviations, over the expectation from the background-only hypothesis. The corresponding expected significance from the standard model for a Higgs boson mass of 125.09 GeV is 4.2 standard deviations. The combined best fit signal strength normalized to the standard model prediction is 1.26^(+0.31)_(−0.26).

Auto-encoders (AEs) have the potential to be effective and generic tools for new physics searches at colliders, requiring little to no model-dependent assumptions. New hypothetical physics signals can be considered anomalies that deviate from the well-known background processes generally expected to describe the whole dataset. We present a search formulated as an anomaly detection (AD) problem, using an AE to define a criterion to decide about the physics nature of an event. In this work, we perform an AD search for manifestations of a dark version of strong force using raw detector images, which are large and very sparse, without leveraging any physics-based pre-processing or assumption on the signals. We propose a dual-encoder design which can learn a compact latent space through conditioning. In the context of multiple AD metrics, we present a clear improvement over competitive baselines and prior approaches. It is the first time that an AE is shown to exhibit excellent discrimination against multiple dark shower models, illustrating the suitability of this method as a performant, model-independent algorithm to deploy, e.g., in the trigger stage of LHC experiments such as ATLAS and CMS.

The dependence of the spectral shape of produced charged hadrons on the size of a colliding system is discussed using a two-component model. As a result, the system-size hierarchy in spectral shape is observed. Next, a hydrodynamic extension of a two-component model for hadroproduction using recent theoretical calculations is suggested to describe the spectra of charged particles produced in heavy-ion collisions in the full range of transverse momenta ${p}_{T}$. Data from heavy-ion collisions measured at the Relativistic Heavy Ion Collider and the Large Hadron Collider are analyzed using the introduced approach and are combined in terms of energy density. The observed regularities might be explained by the formation of a quark-gluon plasma during the collision.

Transverse momentum spectra of charged particle production in heavy-ion collisions are considered in terms of a recently introduced Two Component parameterization combining exponential ("soft") and power-law ("hard") functional forms. The charged hadron densities calculated separately for them are plotted versus number of participating nucleons, $N_{part}$. The obtained dependences are discussed and the possible link between the two component parameterization introduced by the authors and the two component model historically used for the case of heavy-ion collisions is established. Next, the variations of the parameters of the introduced approach with the center of mass energy and centrality are studied using the available data from RHIC and LHC experiments. The spectra shapes are found to show universal dependences on $N_{part}$ for all investigated collision energies.

An overview of the recent searches for Higgs boson pair production at the LHC was presented. The searches were based on approximately 140 $\mathrm{fb}^{-1}$ of data collected by the ATLAS and CMS experiments in proton-proton collisions at $\sqrt{s}$ = 13 TeV. With respect to the previous searches, analysis techniques were significantly improved, and new signatures and decays channels were explored.

More than a thousand 8" silicon sensors will be visually inspected to look for anomalies on their surface during the quality control preceding assembly into the High-Granularity Calorimeter for the CMS experiment at CERN. A deep learning-based algorithm that pre-selects potentially anomalous images of the sensor surface in real time has been developed to automate the visual inspection. The anomaly detection is done by an ensemble of independent deep convolutional neural networks: an autoencoder and a classifier. The performance is evaluated on images acquired in production. The pre-selection reduces the number of images requiring human inspection by 85%, with recall of 97%. Data gathered in production can be used for continuous learning to improve the accuracy incrementally.

Confining dark sectors with pseudo-conformal dynamics can produce Soft Unclustered Energy Patterns, or SUEPs, at the Large Hadron Collider: the production of dark quarks in proton-proton collisions leading to a dark shower and the high-multiplicity production of dark hadrons. The final experimental signature is spherically-symmetric energy deposits by an anomalously large number of soft Standard Model particles with a transverse energy of a few hundred MeV. The dominant background for the SUEP search, if it gets produced via gluon-gluon fusion, is multi-jet QCD events. We have developed a deep learning-based Anomaly Detection technique to reject QCD jets and identify any anomalous signature, including SUEP, in real-time in the High-Level Trigger system of the Compact Muon Solenoid experiment at the Large Hadron Collider. A deep convolutional neural autoencoder network has been trained using QCD events by taking transverse energy deposits in the inner tracker, electromagnetic calorimeter, and hadron calorimeter sub-detectors as 3-channel image data. To tackle the biggest challenge of the task, due to the sparse nature of the data: only ~0.5% of the total ~300 k image pixels have non-zero values, a non-standard loss function, the inverse of the so-called Dice Loss, has been exploited. The trained autoencoder with learned spatial features of QCD jets can detect 40% of the SUEP events, with a QCD event mistagging rate as low as 2%. The model inference time has been measured using the Intel CoreTM i5-9600KF processor and found to be ~20 ms, which perfectly satisfies the High-Level Trigger system's latency of O(100) ms. Given the virtue of the unsupervised learning of the autoencoders, the trained model can be applied to any new physics model that predicts an experimental signature anomalous to QCD jets.

Abstract More than a thousand 8″ silicon sensors will be visually inspected to look for anomalies on their surface during the quality control preceding assembly into the High-Granularity Calorimeter for the CMS experiment at CERN. A deep learning-based algorithm that pre-selects potentially anomalous images of the sensor surface in real time was developed to automate the visual inspection. The anomaly detection is done by an ensemble of independent deep convolutional neural networks: an autoencoder and a classifier. The algorithm was deployed and has been continuously running in production, and data gathered were used to evaluate its performance. The pre-selection reduces the number of images requiring human inspection by 85%, with recall of 97%, and saves 15 person-hours per a batch of a hundred sensors. Data gathered in production can be used for continuous learning to improve the accuracy incrementally.

Unsupervised deep learning techniques are widely used to identify anomalous behaviour. The performance of such methods is a product of the amount of training data and the model size. However, the size is often a limiting factor for the deployment on resource-constrained devices. We present a novel procedure based on knowledge distillation for compressing an unsupervised anomaly detection model into a supervised deployable one and we suggest a set of techniques to improve the detection sensitivity. Compressed models perform comparably to their larger counterparts while significantly reducing the size and memory footprint.

Abstract In high energy physics, one of the most important processes for collider data analysis is the comparison of collected and simulated data. Nowadays the state-of-the-art for data generation is in the form of Monte Carlo (MC) generators. However, because of the upcoming high-luminosity upgrade of the Large Hadron Collider (LHC), there will not be enough computational power or time to match the amount of needed simulated data using MC methods. An alternative approach under study is the usage of machine learning generative methods to fulfill that task. Since the most common final-state objects of high-energy proton collisions are hadronic jets, which are collections of particles collimated in a given region of space, this work aims to develop a convolutional variational autoencoder (ConVAE) for the generation of particle-based LHC hadronic jets. Given the ConVAE’s limitations, a normalizing flow (NF) network is coupled to it in a two-step training process, which shows improvements on the results for the generated jets. The ConVAE+NF network is capable of generating a jet in <?CDATA $18.30 \pm 0.04\,\,{\mu\text{s}}$?> <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" overflow="scroll"> <mml:mn>18.30</mml:mn> <mml:mo>±</mml:mo> <mml:mn>0.04</mml:mn> <mml:mrow> <mml:mi>μ</mml:mi> <mml:mtext>s</mml:mtext> </mml:mrow> </mml:math> , making it one of the fastest methods for this task up to now.

In high energy physics, one of the most important processes for collider data analysis is the comparison of collected and simulated data. Nowadays the state-of-the-art for data generation is in the form of Monte Carlo (MC) generators. However, because of the upcoming high-luminosity upgrade of the LHC, there will not be enough computational power or time to match the amount of needed simulated data using MC methods. An alternative approach under study is the usage of machine learning generative methods to fulfill that task.Since the most common final-state objects of high-energy proton collisions are hadronic jets, which are collections of particles collimated in a given region of space, this work aims to develop a convolutional variational autoencoder (ConVAE) for the generation of particle-based LHC hadronic jets. Given the ConVAE's limitations, a normalizing flow (NF) network is coupled to it in a two-step training process, which shows improvements on the results for the generated jets. The ConVAE+NF network is capable of generating a jet in $18.30 \pm 0.04 \ \mu$s, making it one of the fastest methods for this task up to now.

Charged particles production in heavy-ion collisions is considered in this paper in terms of a recently introduced Two-Component Model. The variations of the parameters of the introduced approach with the center of mass energy and centrality are studied using the available data from RHIC and LHC experiments. The spectra shapes are found to show an universal dependence on the number of participating nucleons $N_{part}$ for all investigated collision energies. Next, the dependences of the obtained charged hadron densities on the $N_{part}$ are discussed. The Two-Component Model allows to separate the charged hadron densities originating from two distinct mechanisms of hadroproduction. Scaling of soft and hard contributions to the spectra with centrality is discussed.

The spectra shape of produced charged hadrons on the size of a colliding system is discussed using a two component model. The hierarchy by the system-size in the spectra shape is observed. Next, the hydrodynamic extension of the model is suggested to describe the spectra of charged particles produced in heavy-ion collisions in the full range of transverse momenta, pT. Data from heavy-ion collisions measured at RHIC and LHC are analyzed and combined in terms of energy density.The observed regularities might be explained by the formation of QGP during the collision.

A search for the standard model Higgs boson is presented in the Vector Boson Fusion production channel with decay to bottom quarks with the CMS experiment at the CERN LHC. A data sample comprising 2.3 fb$^{-1}$ of proton-proton collision at $\sqrt{s}$ = 13 TeV collected during the 2015 running period has been analyzed. Production upper limits at 95\% Confidence Level are derived for a Higgs boson mass of 125 GeV, as well as the fitted signal strength relative to the expectation for the standard model Higgs boson. Results are also combined with the ones obtained with Run1 $\sqrt{s}$ = 8 TeV data collected in 2012.

The shapes of invariant differential cross section for charged particle production as function of transverse momentum measured in heavy-ion collisions are analyzed. The data measured at RHIC and LHC are treated as function of energy density according to a recent theoretical approach. The Boltzmann-like statistical distribution is extracted from the whole statistical ensemble of produced hadrons using the introduced model. Variation of the temperature, characterizing this exponential distribution, is studied as function of energy density.

Total and differential cross sections of vector bosons produced in association with jets are studied at center-of-mass energies √ s = 8 and 13 TeV with the CMS experiment at the LHC.Differential distributions as function of a broad range of kinematic observables are measured and compared to the theoretical predictions.Final states with a vector boson and jets can also be used to study electroweak initiated processes, such as production of the Z boson accompanied by a pair of energetic jets with large invariant mass.The cross section of this electroweak process is measured and the additional hadronic activity of events in a signal-enriched region is studied within the expected rapidity gap region.

Charged particles production in heavy-ions collisions is considered in this paper in terms of a recently introduced Two Component Model.The variations of the parameters of the introduced approach with the center of mass energy and centrality are studied using the available data from RHIC and LHC experiments.The spectra shapes are found to show an universal dependence on the number of participating nucleons N part for all investigated collision energies.Next, the dependences of the obtained charged hadron densities on the N part are discussed.The Two Component Model allows to separate the charged hadron densities originating from two distinct mechanisms of hadroproduction.Scaling of "soft" and "hard" contributions to the spectra with centrality is discussed.

We study how to use Deep Variational Autoencoders for a fast simulation of jets of particles at the LHC. We represent jets as a list of constituents, characterized by their momenta. Starting from a simulation of the jet before detector effects, we train a Deep Variational Autoencoder to return the corresponding list of constituents after detection. Doing so, we bypass both the time-consuming detector simulation and the collision reconstruction steps of a traditional processing chain, speeding up significantly the events generation workflow. Through model optimization and hyperparameter tuning, we achieve state-of-the-art precision on the jet four-momentum, while providing an accurate description of the constituents momenta, and an inference time comparable to that of a rule-based fast simulation.

We present the current stage of research progress towards a one-pass, completely Machine Learning (ML) based imaging calorimeter reconstruction. The model used is based on Graph Neural Networks (GNNs) and directly analyzes the hits in each HGCAL endcap. The ML algorithm is trained to predict clusters of hits originating from the same incident particle by labeling the hits with the same cluster index. We impose simple criteria to assess whether the hits associated as a cluster by the prediction are matched to those hits resulting from any particular individual incident particles. The algorithm is studied by simulating two tau leptons in each of the two HGCAL endcaps, where each tau may decay according to its measured standard model branching probabilities. The simulation includes the material interaction of the tau decay products which may create additional particles incident upon the calorimeter. Using this varied multiparticle environment we can investigate the application of this reconstruction technique and begin to characterize energy containment and performance.

Jets at generator and reconstruction level saved in .npy format. Each jet is represented as an array of jet constituents characterized by their particle momentum in Cartesian coordinates, i.e., (px, py, pz). For both generator-level and reconstruction-level jets, jet constituents are ordered by decreasing pT. The shape of the datasets is [N, 50, 3], where N is the total number of jets, 50 is the number of particles per jet, and 3 is the number of particle features (in order): [px, py, pz].<br> About 1.7M jets split into training, validation and testing sets at 60%, 20% and 20% respectively.

Jets at generator and reconstruction level saved in .npy format. Each jet is represented as an array of jet constituents characterized by their particle momentum in Cartesian coordinates, i.e., (px, py, pz). For both generator-level and reconstruction-level jets, jet constituents are ordered by decreasing pT. The shape of the datasets is [N, 50, 3], where N is the total number of jets, 50 is the number of particles per jet, and 3 is the number of particle features (in order): [px, py, pz].<br> About 1.7M jets split into training, validation and testing sets at 60%, 20% and 20% respectively.

A search for the standard model Higgs boson produced by vector boson fusion in the fully hadronic four-jet topology is presented. The analysis is based on 2.3 fb$^{-1}$ of proton-proton collision data at $\sqrt{s}$ = 13 TeV collected by CMS in 2015. Upper limits, at 95% confidence level, on the production cross section times branching fraction of the Higgs boson decaying to bottom quarks, are derived for a Higgs boson mass of 125 GeV. The fitted signal strength relative to the expectation for the standard model Higgs boson is obtained. Results are also combined with the ones obtained with Run1 data at $\sqrt{s}$ = 8 TeV collected in 2012.

The latest measurements of the vector boson fusion (VBF) and vector boson scattering (VBS) processes from the CMS experiment are presented, using 36 fb$^{-1}$ data collected in proton-proton collisions at $\sqrt{s}$ = 13 TeV at the LHC. The measured total cross sections are compared to the theoretical predictions and limits are set on the presence of anomalous triple and quartic gauge couplings.

We investigate how to improve new physics detection strategies exploiting variational autoencoders and normalizing flows for anomaly detection at the Large Hadron Collider. As a working example, we consider the DarkMachines challenge dataset. We show how different design choices (e.g., event representations, anomaly score definitions, network architectures) affect the result on specific benchmark new physics models. Once a baseline is established, we discuss how to improve the anomaly detection accuracy by exploiting normalizing flow layers in the latent space of the variational autoencoder.