aiqbench analysis
This research introduces hybrid techniques for neural simulation-based inference that significantly lower computational costs while maintaining reliability guarantees comparable to parametric methods. By enabling more efficient offline analyses and paving the way for trigger-level applications, these advances directly address the scalability challenges inherent in applying advanced machine learning models to complex scientific datasets like those in particle physics.
Hybrid Neural Simulation-Based Inference for Robust Applications and Limited-Budget Scenarios
arXiv:2609.36044v1 Announce Type: cross Abstract: We develop two hybrid techniques that approach the performance of neural simulation-based inference (NSBI) analyses while substantially reducing the computational cost of inference and preserving some or all of the reliability guarantees of parametric methods. The first approach is broadly applicable, while the second is tailored to a class of particle physics analyses that admit a semi-parametric NSBI formulation. With only a modest compromise in raw sensitivity, these methods represent an important step toward computationally efficient NSBI in offline analyses and also open the door to the exploration of trigger-level applications in the future. Based on our comparison studies, we recommend the use of our first approach, Latent Categories, for robust and efficient inference.
Abstract:We develop two hybrid techniques that approach the performance of neural simulation-based inference (NSBI) analyses while substantially reducing the computational cost of inference and preserving some or all of the reliability guarantees of parametric methods. The first approach is broadly applicable, while the second is tailored to a class of particle physics analyses that admit a semi-parametric NSBI formulation. With only a modest compromise in raw sensitivity, these methods represent an important step toward computationally efficient NSBI in offline analyses and also open the door to the exploration of trigger-level applications in the future. Based on our comparison studies, we recommend the use of our first approach, Latent Categories, for robust and efficient inference.
High Energy Physics - Phenomenology (hep-ph); Machine Learning (cs.LG); High Energy Physics - Experiment (hep-ex); Data Analysis, Statistics and Probability (physics.data-an)
arXiv:2609.36044 [hep-ph]
(or arXiv:2609.36044v1 [hep-ph] for this version)
https://doi.org/10.48550/arXiv.2609.36044 arXiv-issued DOI via DataCite (pending registration)
Full text under the publisher's open licence · 150 words
Related coverage
News
Quantum Computing Report · Jun 7, 2026
ANCHOR: Error-Controlled Adaptive Numerical Correction for Neural Operator Time Marching
arXiv · Jun 16, 2026
NVIDIA charts a new path for enterprise quantum with what works.
Quantum Zeitgeist · Sep 15, 2026
News
Quantum Computing Report · Jul 25, 2026