Saved in:
Bibliographic Details
Main Authors: Birk, Joschka, Hallin, Anna, Kasieczka, Gregor, Madzharova, Nikol, Pang, Ian, Shih, David
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.04149
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909943068622848
author Birk, Joschka
Hallin, Anna
Kasieczka, Gregor
Madzharova, Nikol
Pang, Ian
Shih, David
author_facet Birk, Joschka
Hallin, Anna
Kasieczka, Gregor
Madzharova, Nikol
Pang, Ian
Shih, David
contents Next token prediction is an attractive pre-training task for jet foundation models, in that it is simulation free and enables excellent generative capabilities that can transfer across datasets. Here we study multiple improvements to next token prediction, building on the initial work of OmniJet-$α$. Instead of tokenizing particles and subsequently only using the token-ID as the model input for both the generative and the classification task, we adopt a hybrid setup, which allows us to use continuous feature vectors as model input while only using token-IDs in the next token prediction target. Secondly, we explore a combined pre-training strategy that combines masked particle modeling and generative learning objectives. Taken together, these changes greatly improve the performance in downstream classification tasks without any loss in generative performance.
format Preprint
id arxiv_https___arxiv_org_abs_2512_04149
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Enhancing next token prediction based pre-training for jet foundation models
Birk, Joschka
Hallin, Anna
Kasieczka, Gregor
Madzharova, Nikol
Pang, Ian
Shih, David
High Energy Physics - Phenomenology
Machine Learning
High Energy Physics - Experiment
Data Analysis, Statistics and Probability
Next token prediction is an attractive pre-training task for jet foundation models, in that it is simulation free and enables excellent generative capabilities that can transfer across datasets. Here we study multiple improvements to next token prediction, building on the initial work of OmniJet-$α$. Instead of tokenizing particles and subsequently only using the token-ID as the model input for both the generative and the classification task, we adopt a hybrid setup, which allows us to use continuous feature vectors as model input while only using token-IDs in the next token prediction target. Secondly, we explore a combined pre-training strategy that combines masked particle modeling and generative learning objectives. Taken together, these changes greatly improve the performance in downstream classification tasks without any loss in generative performance.
title Enhancing next token prediction based pre-training for jet foundation models
topic High Energy Physics - Phenomenology
Machine Learning
High Energy Physics - Experiment
Data Analysis, Statistics and Probability
url https://arxiv.org/abs/2512.04149