14–18 Sept 2026
Europe/Vienna timezone

LI-JEPA: Self-supervised pretraining of Lorentz Invariant Models

16 Sept 2026, 10:30
20m

Speaker

Andreas Hermansen (Universite de Geneve (CH))

Description

Deep learning classifiers for high-energy physics have recently advanced along two largely separate tracks: self-supervised pretraining on large unlabelled datasets, and architectures that respect Lorentz symmetry by construction. How these directions combine has not been shown.

To address this, we integrate Lorentz-invariant models into the LeJEPA pretraining framework, in which the network is trained to produce a latent representation that is invariant under predefined augmentations. This naturally aligns with the philosophy of Lorentz-invariant models, where invariance under augmentations consisting purely of a Lorentz transformation is trivially enforced by the network.

We evaluate on JetClass under matched pretraining data, augmentations, and optimization, varying only the backbone. Under this comparison the Lorentz-invariant model achieves substantially superior linear separability and downstream tagging accuracy relative to a non-invariant scalar transformer, showing that exact physical symmetries and self-supervised pretraining are complementary rather than competing formulations.

Author

Andreas Hermansen (Universite de Geneve (CH))

Co-author

Tobias Golling (Universite de Geneve (CH))

Presentation materials

There are no materials yet.