ATLAS Jet Flavor Tagging with AI: The GN2 Algorithm
he ATLAS Experiment at CERN has embraced modern AI techniques to revolutionise jet flavour tagging, a crucial process in analysing particle collisions. A new algorithm called GN2, powered by a Transformer neural network, directly analyses information from particle tracks and jets, eliminating the need for previous, hand-crafted algorithms. This advancement significantly improves the identification of b-jets and c-jets, which are vital for Standard Model measurements and the search for new physics phenomena. The ATLAS Collaboration has established robust pipelines to integrate and train these AI algorithms, leading to a substantial leap in performance and offering deeper insights into the physics signatures learned by the model. This innovative approach is already having a significant impact on ATLAS physics research, including enhancing the precision of Higgs boson studies and the search for new particles.Paper link: https://arxiv.org/pdf/2505.19689
--------
13:50
--------
13:50
Supersymmetry Search at CMS Experiment: Boosted Objects and Razor Variables
The CMS Collaboration's "An all-round boosted chase for supersymmetry" explores the concept of supersymmetry (SUSY), a theoretical extension of the Standard Model of particle physics. The article details how scientists at CERN's CMS Experiment are searching for superparticles by analysing final states that include Lorentz-boosted objects, which are formed when heavy superparticles decay into lighter, high-momentum particles. This new "razor boost" analysis broadens the search to 25 distinct final states, including boosted W, Z, Higgs bosons, top quarks, and uniquely, boosted leptonic jets, using machine learning and razor kinematic variables to identify potential SUSY signals. Although no significant deviations from Standard Model predictions were observed, the findings have allowed the collaboration to set strong limits on superparticle production rates and masses within various SUSY models, with further interpretations of complete SUSY models ongoing.Links: https://cms-results.web.cern.ch/cms-results/public-results/preliminary-results/SUS-23-014/index.html
--------
12:50
--------
12:50
Jet Intercalibration at the ATLAS Experiment: Uncertainty and Consistency Evaluation
The current episode discusses a research paper from the ATLAS Collaboration concerning the evaluation of statistical uncertainties in the jet η-intercalibration method for the ATLAS experiment at CERN's Large Hadron Collider. The paper, identified as ATL-PHYS-PUB-2025-027, explains a new approach utilising the bootstrap method to determine calibration factor uncertainties and their correlations more accurately. This innovative technique is compared against previous methods, with figures illustrating distributions, uncertainty ratios, and correlation matrices. Additionally, the paper introduces a method for quantifying the statistical compatibility of jet observables within the calibration procedure, employing a pull statistic to assess robustness.Links: https://cds.cern.ch/record/2935135/files/ATL-PHYS-PUB-2025-027.pdf
--------
14:22
--------
14:22
Sharper Sight for Electron-Positron Pairs at the CMS Experiment
In this episode, we discuss the details of the CMS experiment's innovative advancement in particle detection, specifically focusing on its enhanced ability to identify highly collimated electron-positron pairs. Previously, the CMS detector struggled to differentiate these pairs when they travelled too closely, often registering them as a single particle. To overcome this, the CMS Collaboration developed a new machine learning-based technique that significantly improves the detector's resolution, enabling it to distinguish pairs with extremely small angular separations. The text explains how this new method has been rigorously tested and validated using both simulations and real-world data, confirming its efficacy in energy measurement and consistency. This improved capability will allow CMS to conduct more precise searches for new phenomena beyond the Standard Model, particularly theories predicting the existence of lightweight bosons that decay into such electron-positron pairs.Read more about it in: https://cms-results.web.cern.ch/cms-results/public-results/preliminary-results/EGM-24-002/index.html
--------
19:22
--------
19:22
Rare Higgs decays investigated by the ATLAS experiment
This episode details the ATLAS Collaboration's ongoing research into rare Higgs boson decays at the Large Hadron Collider (LHC), specifically focusing on decays to muons (H→μμ) and to a Z boson and a photon (H→Zγ). These studies, utilising data from LHC Run 3 (2022–2024) and combined with Run 2 (2015–2018) data, aim to test the Standard Model by precisely measuring these infrequent processes. The documents explain the experimental methods used to identify these rare events, including sophisticated data analysis techniques and event categorisation, and present the statistical significance of the observed evidence. The findings demonstrate improved sensitivity and contribute to a more comprehensive understanding of the Higgs boson's properties.Papers:https://arxiv.org/pdf/2507.03595https://cds.cern.ch/record/2937635/files/ATLAS-CONF-2025-007.pdf
Go beyond the headlines and dive into the real science of the Large Hadron Collider. Each episode, your AI particle physicists Bob and Alice break down the latest, most significant papers from the LHC experiments at CERN. From new measurements of the Higgs boson to the ongoing search for physics beyond the Standard Model, this podcast offers a clear, expert review of cutting-edge research's data, methods, and implications. Perfect for physics students, researchers, and anyone who wants to understand the universe at its most fundamental level.
España