ν-Flows: Conditional Neutrino Regression

Abstract

We present $\nu$-Flows, a novel method for restricting the likelihood space of neutrino kinematics in high energy collider experiments using conditional normalizing flows and deep invertible neural networks. This method allows the recovery of the full neutrino momentum which is usually left as a free parameter and permits one to sample neutrino values under a learned conditional likelihood given event observations. We demonstrate the success of $\nu$-Flows in a case study by applying it to simulated semileptonic $t\bar{t}$ events and show that it can lead to more accurate momentum reconstruction, particularly of the longitudinal coordinate. We also show that this has direct benefits in a downstream task of jet association, leading to an improvement of up to a factor of 1.41 compared to conventional methods.

26 pages, 15 figures

References in corpus (6)

• The automated computation of tree-level and next-to-leading order differential cross sections, and their matching to parton shower simulations
• LHAPDF6: parton density access in the LHC precision era
• Automatic spin-entangled decays of heavy resonances in Monte Carlo simulations
• Guided Image Generation with Conditional Invertible Neural Networks
• Phase Space Sampling and Inference from Weighted Events with Autoregressive Flows
• CURTAINs for your Sliding Window: Constructing Unobserved Regions by Transforming Adjacent Intervals

Cited by in corpus (26)

• MadNIS -- Neural Multi-Channel Importance Sampling
• Two Invertible Networks for the Matrix Element Method
• Quantum Information meets High-Energy Physics: Input to the update of the European Strategy for Particle Physics
• The MadNIS Reloaded
• Returning CP-Observables to The Frames They Belong
• CaloFlow for CaloChallenge Dataset 1
• CaloDREAM -- Detector Response Emulation via Attentive flow Matching
• The Landscape of Unfolding with Machine Learning
• ${\nu }^2$-Flows: Fast and improved neutrino reconstruction in multi-neutrino final states with conditional normalizing flows
• Precision-Machine Learning for the Matrix Element Method
• Normalizing Flows for High-Dimensional Detector Simulations
• Calorimeter shower superresolution
• Differentiable MadNIS-Lite
• Reconstruction of Unstable Heavy Particles Using Deep Symmetry-Preserving Attention Networks
• CaloChallenge 2022: A Community Challenge for Fast Calorimeter Simulation
• Anomaly detection with flow-based fast calorimeter simulators
• Accurate Surrogate Amplitudes with Calibrated Uncertainties
• Unifying Simulation and Inference with Normalizing Flows
• Machine Learning-Assisted Unfolding for Neutrino Cross-section Measurements with the OmniFold Technique
• Angular correlations in associated top-quark and dark matter production at Large Hadron Collider
• Extrapolating Jet Radiation with Autoregressive Transformers
• Tools for Unbinned Unfolding
• Observable Optimization for Precision Theory: Machine Learning Energy Correlators
• How to Unfold Top Decays
• Machine learning in top quark physics at ATLAS and CMS
• Variational inference for pile-up removal at hadron colliders with diffusion models