Mind the Gap: Navigating Inference with Optimal Transport Maps
physics.data-an, cs.LG, hep-ex, stat.ML
Submitted: 2025-07-09
Updated: 2026-09-06
Comments: 31 pages, 13 figures
Project page: http://stat.stanford.edu/~owen/mc
License: http://creativecommons.org/licenses/by/4.0/
The gist: Machine learning (ML) techniques have recently enabled enormous gains in sensitivity to new phenomena across the sciences.
Terminology
Abstract
Machine learning (ML) techniques have recently enabled enormous gains in sensitivity to new phenomena across the sciences. In particle physics, much of this progress has relied on excellent simulations of a wide range of physical processes. However, due to the sophistication of modern machine learning algorithms and their reliance on high-quality training samples, discrepancies between simulation and experimental data can significantly limit their effectiveness. In this work, we present a solution to this ``misspecification'' problem: a model calibration approach based on optimal transport, which we apply to high-dimensional simulations for the first time. We demonstrate the performance of our approach through jet tagging, using a dataset inspired by the CMS experiment at the Large Hadron Collider. A 128-dimensional internal jet representation from a powerful general-purpose classifier is studied; after calibrating this internal ``latent'' representation, we find that a wide variety of quantities derived from it for downstream tasks are also properly calibrated: using this calibrated high-dimensional representation, powerful new applications of jet flavor information can be utilized in LHC analyses. This is a key step toward allowing the unbiased use of ``foundation models'' in particle physics. More broadly, this calibration framework has broad applications for correcting high-dimensional simulations across the sciences.
Sources
- Neural Methods for Amortized Inference
- A Comprehensive Guide to Simulation-based Inference in Computational Biology
- Investigating the Impact of Model Misspecification in Neural Simulation-based Inference
- Tests for model misspecification in simulation-based inference: from local distortions to global model checks
- Detecting Model Misspecification in Amortized Bayesian Inference with Neural Networks
- Learning Robust Statistics for Simulation-based Inference under Model Misspecification
- A continuous calibration of the ATLAS flavour-tagging classifiers via optimal transportation maps
- Addressing Misspecification in Simulation-based Inference through Data-driven Calibration
- Optimal transport mapping via input convex neural networks
- Convex Potential Flows: Universal Probability Distributions with Optimal Transport and Convex Optimization
- Do Neural Optimal Transport Solvers Work? A Continuous Wasserstein-2 Benchmark
- Supervised Training of Conditional Monge Maps
- General-purpose event generators for LHC physics
- The ATLAS Simulation Infrastructure
- Transforming jet flavour tagging at ATLAS
- Is Tokenization Needed for Masked Particle Modelling?
- Accelerating Resonance Searches via Signature-Oriented Pre-training
- A reconstruction of the initial conditions of the Universe by optimal mass transportation
- SHAM-OT: Rapid Subhalo Abundance Matching with Optimal Transport
- One flow to correct them all: improving simulations in high-energy physics with a single normalising flow and a switch
Related papers
- Modular Aggregation as a Debiasing Method for Non-Stationary Discrete Sources: Convergence and Numerical Validatio
- Interpreting "Interpretability" and Explaining "Explainability" in Machine Learning in Physics
- Dynamics-Based Intrinsic Signal Model for High-Dimensional, Small-Sample Data
- Exploring new directions in enhancing the ACTS parameter optimization suite