Olmo Cerri

A bstract Using variational autoencoders trained on known physics processes, we develop a one-sided threshold test to isolate previously unseen processes as outlier events. Since the autoencoder training does not depend on any specific new physics signature, the proposed procedure doesn’t make specific assumptions on the nature of new physics. An event selection based on this algorithm would be complementary to classic LHC searches, typically based on model-dependent hypothesis testing. Such an algorithm would deliver a list of anomalous events, that the experimental collaborations could further scrutinize and even release as a catalog, similarly to what is typically done in other scientific domains. Event topologies repeating in this dataset could inspire new-physics model building and new experimental searches. Running in the trigger system of the LHC experiments, such an application could identify anomalous events that would be otherwise lost, extending the scientific reach of the LHC.

Abstract We investigate the performance of a jet identification algorithm based on interaction networks (JEDI-net) to identify all-hadronic decays of high-momentum heavy particles produced at the LHC and distinguish them from ordinary jets originating from the hadronization of quarks and gluons. The jet dynamics are described as a set of one-to-one interactions between the jet constituents. Based on a representation learned from these interactions, the jet is associated to one of the considered categories. Unlike other architectures, the JEDI-net models achieve their performance without special handling of the sparse input jet representation, extensive pre-processing, particle ordering, or specific assumptions regarding the underlying detector geometry. The presented models give better results with less model parameters, offering interesting prospects for LHC applications.

Abstract We apply an Adversarially Learned Anomaly Detection (ALAD) algorithm to the problem of detecting new physics processes in proton–proton collisions at the Large Hadron Collider. Anomaly detection based on ALAD matches performances reached by Variational Autoencoders, with a substantial improvement in some cases. Training the ALAD algorithm on 4.4 fb $$^{-1}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mrow /> <mml:mrow> <mml:mo>-</mml:mo> <mml:mn>1</mml:mn> </mml:mrow> </mml:msup> </mml:math> of 8 TeV CMS Open Data, we show how a data-driven anomaly detection and characterization would work in real life, re-discovering the top quark by identifying the main features of the $$t \bar{t}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>t</mml:mi> <mml:mover> <mml:mrow> <mml:mi>t</mml:mi> </mml:mrow> <mml:mrow> <mml:mo>¯</mml:mo> </mml:mrow> </mml:mover> </mml:mrow> </mml:math> experimental signature at the LHC.

At the Large Hadron Collider, the high-transverse-momentum events studied by experimental collaborations occur in coincidence with parasitic low-transverse-momentum collisions, usually referred to as pileup. Pileup mitigation is a key ingredient of the online and offline event reconstruction as pileup affects the reconstruction accuracy of many physics observables. We present a classifier based on Graph Neural Networks, trained to retain particles coming from high-transverse-momentum collisions, while rejecting those coming from pileup collisions. This model is designed as a refinement of the PUPPI algorithm (D. Bertolini et al., JHEP 10, 059 (2014)), employed in many LHC data analyses since 2015. Thanks to an extended basis of input information and the learning capabilities of the considered network architecture, we show an improvement in pileup-rejection performances with respect to state-of-the-art solutions.

We develop an algorithm based on an interaction network to identify high-transverse-momentum Higgs bosons decaying to bottom quark-antiquark pairs and distinguish them from ordinary jets that reflect the configurations of quarks and gluons at short distances. The algorithm's inputs are features of the reconstructed charged particles in a jet and the secondary vertices associated with them. Describing the jet shower as a combination of particle-to-particle and particle-to-vertex interactions, the model is trained to learn a jet representation on which the classification problem is optimized. The algorithm is trained on simulated samples of realistic LHC collisions, released by the CMS Collaboration on the CERN Open Data Portal. The interaction network achieves a drastic improvement in the identification performance with respect to state-of-the-art algorithms.

We show how an event topology classification based on deep learning could be used to improve the purity of data samples selected in real time at the Large Hadron Collider. We consider different data representations, on which different kinds of multi-class classifiers are trained. Both raw data and high-level features are utilized. In the considered examples, a filter based on the classifier’s score can be trained to retain $$\sim 99\%$$ of the interesting events and reduce the false-positive rate by more than one order of magnitude. By operating such a filter as part of the online event selection infrastructure of the LHC experiments, one could benefit from a more flexible and inclusive selection strategy while reducing the amount of downstream resources wasted in processing false positives. The saved resources could translate into a reduction of the detector operation cost or into an effective increase of storage and processing capabilities, which could be reinvested to extend the physics reach of the LHC experiments.

Deterministic generation of single photons is essential for many quantum information technologies. A bulk optical nonlinearity emitting a photon pair, where the measurement of one of the photons heralds the presence of the other, is commonly used with the caveat that the single-photon emission rate is constrained due to a trade-off between multiphoton events and pair emission rate. Using an efficient and low noise photon-number-resolving superconducting nanowire detector we herald, in real time, a single photon at telecommunication wavelength. We perform a second-order photon correlation ${g}^{2}(0)$ measurement of the signal mode conditioned on the measured photon number of the idler mode for various pump powers and demonstrate an improvement of a heralded single-photon source. We develop an analytical model using a phase-space formalism that encompasses all multiphoton effects and relevant imperfections, such as loss and multiple Schmidt modes. We perform a maximum-likelihood fit to test the agreement of the model to the data and extract the best-fit mean photon number $\ensuremath{\mu}$ of the pair source for each pump power. A maximum reduction of $0.118\ifmmode\pm\else\textpm\fi{}0.012$ in the photon ${g}^{2}(0)$ correlation function at $\ensuremath{\mu}=0.327\ifmmode\pm\else\textpm\fi{}0.007$ is obtained, indicating a strong suppression of multiphoton emissions. For a fixed ${g}^{2}(0)=7\ifmmode\times\else\texttimes\fi{}{10}^{\ensuremath{-}3}$, we increase the single pair generation probability by 25%. Our experiment, built using fiber-coupled and off-the-shelf components, delineates a path to engineering ideal sources of single photons.

We present a fast-simulation application based on a deep neural network, designed to create large analysis-specific datasets. Taking as an example the generation of W + jet events produced in s= 13 TeV proton-proton collisions, we train a neural network to model detector resolution effects as a transfer function acting on an analysis-specific set of relevant features, computed at generation level, i.e., in absence of detector effects. Based on this model, we propose a novel fast-simulation workflow that starts from a large amount of generator-level events to deliver large analysis-specific samples. The adoption of this approach would result in about an order-of-magnitude reduction in computing and storage requirements for the collision simulation workflow. This strategy could help the high energy physics community to face the computing challenges of the future High-Luminosity LHC.

A bstract We study the impact of precision timing detection systems on the LHC experiments’ long-lived particle search program during the HL-LHC era. We develop algorithms that allow us to reconstruct the mass of such charged particles and perform particle identification using the time-of-flight measurement. We investigate the reach for benchmark scenarios as a function of the timing resolution, and find sensitivity improvement of up to a factor of ten over searches that use ionization energy loss information, depending on the particle’s mass.

We study the expected sensitivity to measure the branching ratio of Higgs boson decays to invisible particles at a future circular $$e^+e^-$$ collider (FCC-ee) in the process $$e^+e^-\rightarrow HZ$$ with $$Z\rightarrow \ell ^+\ell ^-$$ ( $$\ell =e$$ or $$\mu $$ ) using an integrated luminosity of 3.5 ab $$^{-1}$$ at a center-of-mass energy $$\sqrt{s}=240$$ GeV. The impact of the energy spread of the FCC-ee beam and of the resolution in the reconstruction of the leptons is discussed. The minimum branching ratio for a $$5\sigma $$ observation after 3.5 ab $$^{-1}$$ of data taking is $$1.7\pm 0.1\%(stat+syst) $$ . The branching ratio exclusion limit at 95% CL is $$0.63 \pm 0.22\%((stat+syst))$$ .

$W$ bosons are produced at LHC from a forward-backward symmetric initial state. Their decay to a charged lepton and a neutrino has a strong spin analysing power. The combination of these effects results in characteristic distributions of the pseudorapidity of the leptons decaying from $W^+$ and $W^-$ of different helicity. This observation may open the possibility to measure precisely the $W^+$ and $W^-$ rapidity distributions for the two transverse polarisation states of $W$ bosons produced at small transverse momentum.

We discuss a model-independent strategy for boosting new physics searches with the help of an unsupervised anomaly detection algorithm. Prior to a search, each input event is preprocessed by the algorithm - a variational autoencoder (VAE). Based on the loss assigned to each event, input data can be split into a background control sample and a signal enriched sample. Following this strategy, one can enhance the sensitivity to new physics with no assumption on the underlying new physics signature. Our results show that a typical BSM search on the signal enriched group is more sensitive than an equivalent search on the original dataset.

The MIP Timing Detector will provide additional timing capabilities for detection of minimum ionizing particles (MIPs) at CMS during the High Luminosity LHC era, improving event reconstruction and pileup rejection. The central portion of the detector, the Barrel Timing Layer (BTL), will be instrumented with LYSO:Ce crystals and Silicon Photomultipliers (SiPMs) providing a time resolution of about 30 ps at the beginning of operation, and degrading to 50-60 ps at the end of the detector lifetime as a result of radiation damage. In this work, we present the results obtained using a 120 GeV proton beam at the Fermilab Test Beam Facility to measure the time resolution of unirradiated sensors. A proof-of-concept of the sensor layout proposed for the barrel region of the MTD, consisting of elongated crystal bars with dimensions of about 3 x 3 x 57 mm$^3$ and with double-ended SiPM readout, is demonstrated. This design provides a robust time measurement independent of the impact point of the MIP along the crystal bar. We tested LYSO:Ce bars of different thickness (2, 3, 4 mm) with a geometry close to the reference design and coupled to SiPMs manufactured by Hamamatsu and Fondazione Bruno Kessler. The various aspects influencing the timing performance such as the crystal thickness, properties of the SiPMs (e.g. photon detection efficiency), and impact angle of the MIP are studied. A time resolution of about 28 ps is measured for MIPs crossing a 3 mm thick crystal bar, corresponding to an MPV energy deposition of 2.6 MeV, and of 22 ps for the 4.2 MeV MPV energy deposition expected in the BTL, matching the detector performance target for unirradiated devices.

We apply an Adversarially Learned Anomaly Detection (ALAD) algorithm to the problem of detecting new physics processes in proton-proton collisions at the Large Hadron Collider. Anomaly detection based on ALAD matches performances reached by Variational Autoencoders, with a substantial improvement in some cases. Training the ALAD algorithm on 4.4 fb-1 of 8 TeV CMS Open Data, we show how a data-driven anomaly detection and characterization would work in real life, re-discovering the top quark by identifying the main features of the t-tbar experimental signature at the LHC.

At the Large Hadron Collider, the high transverse-momentum events studied by experimental collaborations occur in coincidence with parasitic low transverse-momentum collisions, usually referred to as pileup. Pileup mitigation is a key ingredient of the online and offline event reconstruction as pileup affects the reconstruction accuracy of many physics observables. We present a classifier based on Graph Neural Networks, trained to retain particles coming from high-transverse-momentum collisions, while rejecting those coming from pileup collisions. This model is designed as a refinement of the PUPPI algorithm, employed in many LHC data analyses since 2015. Thanks to an extended basis of input information and the learning capabilities of the considered network architecture, we show an improvement in pileup-rejection performances with respect to state-of-the-art solutions.

\(h^\pm \to \tau^\pm \nu\) signal events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(LQ \to b \tau\) signal events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(A \to 4\ell\) signal events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(h^0 \to \tau^+ \tau^-\) signal events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

At the Large Hadron Collider, the high transverse-momentum events studied by experimental collaborations occur in coincidence with parasitic low transverse-momentum collisions, usually referred to as pileup. Pileup mitigation is a key ingredient of the online and offline event reconstruction as pileup affects the reconstruction accuracy of many physics observables. We present a classifier based on Graph Neural Networks, trained to retain particles coming from high-transverse-momentum collisions, while rejecting those coming from pileup collisions. This model is designed as a refinement of the PUPPI algorithm, employed in many LHC data analyses since 2015. Thanks to an extended basis of input information and the learning capabilities of the considered network architecture, we show an improvement in pileup-rejection performances with respect to state-of-the-art solutions.

We present a fast simulation application based on a Deep Neural Network, designed to create large analysis-specific datasets. Taking as an example the generation of W+jet events produced in sqrt(s)= 13 TeV proton-proton collisions, we train a neural network to model detector resolution effects as a transfer function acting on an analysis-specific set of relevant features, computed at generation level, i.e., in absence of detector effects. Based on this model, we propose a novel fast-simulation workflow that starts from a large amount of generator-level events to deliver large analysis-specific samples. The adoption of this approach would result in about an order-of-magnitude reduction in computing and storage requirements for the collision simulation workflow. This strategy could help the high energy physics community to face the computing challenges of the future High-Luminosity LHC.

\(t \bar t\) background events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

QCD multijet background events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(Z \to \ell \ell\) background events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(W \to \ell \nu\) background events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

This Technical Memorandum (TM) summarizes the Fermilab Test Beam operations for FY2017. It is one of a series of annual publications intended to gather information in one place. In this case, the information concerns the individual experiments that ran at FTBF and are listed in Table 1. Each experiment section was prepared by the relevant authors, and was edited for inclusion in this summary.

As part of the Phase II upgrade program, the Compact Muon Solenoid (CMS) detector will incorporate a new timing layer designed to measure minimum ionizing particles (MIPs) with a time resolution of $\sim$30 ps. Precision timing will mitigate the impact of the challenging levels of pileup expected at the High Luminosity LHC. The time information assigned to each track will enable the use of 4D-vertexing which will render a 5-fold pile-up reduction, thus recovering the current conditions. Precision timing will also enable new time-based isolations and improved $b$-tagging algorithms. All of this translates into a $\sim20\%$ gain in effective luminosity when looking at di-Higgs boson events decaying to a pair of $b$-quarks and two photons. We present the expected improvements in physics performance with precision timing with the upgraded CMS detector.

Deep learning plays a significant role in jet tagging. Interaction network / message passing network are parameter efficient. The proposed network out-performs some other deep learning approaches. There is promising direction for future taggers and other problems.

Topology Classification Dataset: raw images of simulated proton-proton LHC collisions at 13 TeV Full description at https://arxiv.org/abs/1807.00083

This Technical Memorandum (TM) summarizes the Fermilab Test Beam Faciltiy (FTBF) operations for FY2019. It is one of a series of annual publications intended to gather information in one place. This TM discusses the experiments performed at the Test Beam from November 2018 to July 2019. The experiments are listed in Table 1. Each experiment wrote a summary that was edited for clarity and is included in this report.

We apply an Adversarially Learned Anomaly Detection (ALAD) algorithm to the problem of detecting new physics processes in proton-proton collisions at the Large Hadron Collider. Anomaly detection based on ALAD matches performances reached by Variational Autoencoders, with a substantial improvement in some cases. Training the ALAD algorithm on 4.4 fb-1 of 8 TeV CMS Open Data, we show how a data-driven anomaly detection and characterization would work in real life, re-discovering the top quark by identifying the main features of the t-tbar experimental signature at the LHC.

\(LQ \to b \tau\) signal events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(Z \to \ell \ell\) background events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(h^0 \to \tau^+ \tau^-\) signal events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(h^\pm \to \tau^\pm \nu\) signal events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

Topology Classification Dataset: raw images of simulated proton-proton LHC collisions at 13 TeV Full description at https://arxiv.org/abs/1807.00083

QCD multijet background events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

\(W \to \ell \nu\) background events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

Topology Classification Dataset: abstract images of simulated proton-proton LHC collisions at 13 TeV Full description at https://arxiv.org/abs/1807.00083

\(A \to 4\ell\) signal events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276

Topology Classification Dataset: abstract images of simulated proton-proton LHC collisions at 13 TeV Full description at https://arxiv.org/abs/1807.00083

\(t \bar t\) background events reconstructed by inclusive single-muon selection. Events are represented as an array of physics-motivated high-level features. Details are given in https://arxiv.org/abs/1811.10276