AI Summary of Scholarly Research

This page presents an AI-generated summary of a published research paper. The original authors did not write or review this article. [See full disclosure ↓]

Transformer identifies strongly modified jets in heavy-ion collisions

Research area:physics-astronomy

What the study found

The study found that a Transformer classifier can detect information in jet data that is not captured by other models using high-level physical observables. It also identified a class of jets that the authors describe as unequivocally modified, even when medium response and underlying event contamination were included.

Why the authors say this matters

The authors say this matters because vacuum-like jets, meaning jets that experienced little interaction with the quark-gluon plasma, dilute the overall observed modification in nucleus-nucleus collision samples. The study suggests that being able to judge jet modification on a jet-by-jet basis would help overcome this limitation.

What the researchers tested

The researchers tested a Transformer classifier trained on a low-level representation of jets using the four-momenta of all jet constituents. They compared its performance with other architectures that use high-level physical observables as input. The study also examined the experimentally relevant case in which medium response and underlying event contamination are both accounted for.

What worked and what didn't

The Transformer was able to capture discriminating information that other tested architectures did not access. In the setting studied, it could identify a class of jets that were clearly modified. The authors also performed a robust estimate of an upper bound on the fraction of nucleus-nucleus collision jets that are indistinguishable from proton-proton jets.

What to keep in mind

The abstract does not provide numerical results in the text available here. It also does not describe detailed limitations beyond the fact that the estimate concerns an upper bound for jets that are, for all purposes, indistinguishable from proton-proton jets.

Key points

  • The study says a Transformer can find jet information not captured by models using high-level observables.
  • It identified a class of jets described as unequivocally modified.
  • The authors say vacuum-like jets dilute the observed modification in nucleus-nucleus collision samples.
  • The researchers estimated an upper bound on jets in nucleus-nucleus collisions that are indistinguishable from proton-proton jets.
  • The analysis included medium response and underlying event contamination.

Disclosure

Research title:
Transformer identifies strongly modified jets in heavy-ion collisions
Authors:
Miguel Crispim Romão, João A. Gonçalves, José Guilherme Milhano
Institutions:
Durham University, University of Lisbon, University of Lisbon
Publication date:
2026-04-27
OpenAlex record:
View
AI provenance: This post was generated by gpt-5.4-mini (OpenAI). The original authors did not write or review this post.