aiqbench analysis
This research addresses a critical bottleneck in applying machine learning to physics by developing adaptive domain adaptation techniques that distinguish between nuisance parameters and genuine physical mismatches. By reweighting simulated events, the method prevents standard adversarial approaches from introducing biases anchored on incorrect simulation priors when transferring models to experimental data. This advancement is essential for improving the reliability of neural networks in astrophysics and high-energy physics where simulation fidelity cannot be assumed.
Safe Domain Adaptation for Physics: Overcoming Nuisances, Label Shifts, and Simulation Priors
arXiv:2608.18190v1 Announce Type: new Abstract: Domain adaptation is widely used to make neural networks trained on simulations applicable to experimental data. Its premise is that the two domains differ only in nuisances, and that the quantity of interest is distributed identically in both. In physics neither assumption holds: simulations can be wrong about the physics, and the distribution of the target quantity - an energy spectrum, a redshift distribution - is often the measurement itself. We study the consequences of such mismatches on a toy air-shower benchmark in which a detector-response nuisance, a physical simulation shift, and an energy-spectrum shift can be switched on separately or together. Standard adversarial adaptation handles the conditional shifts, but once the two spectra differ it aligns them, replacing an uncontrolled bias by one anchored on the simulation prior. We present adaptive domain adaptation, which reweights the simulated events so as to focus domain adaptation on the genuine physical mismatch alone. Since the predicted spectrum depends on model training configuration, we provide a label-free model selection rule for selecting the near-the-best operation point.
Abstract:Domain adaptation is widely used to make neural networks trained on simulations applicable to experimental data. Its premise is that the two domains differ only in nuisances, and that the quantity of interest is distributed identically in both. In physics neither assumption holds: simulations can be wrong about the physics, and the distribution of the target quantity - an energy spectrum, a redshift distribution - is often the measurement itself. We study the consequences of such mismatches on a toy air-shower benchmark in which a detector-response nuisance, a physical simulation shift, and an energy-spectrum shift can be switched on separately or together. Standard adversarial adaptation handles the conditional shifts, but once the two spectra differ it aligns them, replacing an uncontrolled bias by one anchored on the simulation prior. We present adaptive domain adaptation, which reweights the simulated events so as to focus domain adaptation on the genuine physical mismatch alone. Since the predicted spectrum depends on model training configuration, we provide a label-free model selection rule for selecting the near-the-best operation point.
Machine Learning (cs.LG); Instrumentation and Methods for Astrophysics (astro-ph.IM); Data Analysis, Statistics and Probability (physics.data-an)
arXiv:2608.18190 [cs.LG]
(or arXiv:2608.18190v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2608.18190 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 279 words
Related coverage
Biological Superintelligence
LessWrong · Jul 30, 2026
Problems of Proliferation
LessWrong · Sep 29, 2026
News
Quantum Computing Report · Jul 25, 2026
Validated Adaptation for Aerial Crowd Monitoring at Mass Gathering Scale: A Deployment Protocol, a Severity Law, and a Diagnostic for Label-Free Drone Crowd Counting, Toward the FIFA World Cup 2034 (Saudi Arabia)
arXiv · Aug 19, 2026