Metalearning and data augmentation for mass-generalized jet taggers

Dolan, Matthew J.; Ore, Ayodele

doi:10.1103/physrevd.105.094030

Cited by 6 publications

(4 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Indeed, the transverse kinematic imbalance can give an enhanced sensitivity to the Fermi motion of initial state nucleons, the final state interactions and the multinucleon processes. Additionally, the transverse momentum imbalance 𝛿 𝑝 𝑇 could allow to select a sample enriched with antineutrino interactions with hydrogen from the region of low 𝛿 𝑝 𝑇 , which would yield improved constraints on the flux uncertainties [12].…”

Section: Pos(now2022)027mentioning

confidence: 99%

The T2K Near Detector Upgrade

Chakrani¹

2023

Proceedings of Neutrino Oscillation Workshop — PoS(NOW2022)

View full text Add to dashboard Cite

The T2K experiment is a long-baseline neutrino oscillation experiment that aims to precisely measure neutrino oscillation parameters. The muon (anti)neutrino beam produced at J-PARC is measured at the near detector complex and at the far detector Super-Kamiokande. The major role of the near detector ND280 is to constrain systematic uncertainties that affect neutrino oscillation measurements. T2K is currently upgrading ND280 to fully leverage the expected increase of statistics over the upcoming years and to further improve the constraints on those uncertainties, particularly the ones related to the neutrino interaction model. The capabilities of full polar angle acceptance, lower proton tracking threshold as well as reconstruction of neutron kinematics that this upgrade offers will open the door to explore new physics with unprecedented precision.

show abstract

Section: Pos(now2022)027mentioning

confidence: 99%

The T2K Near Detector Upgrade

Chakrani¹

2023

Proceedings of Neutrino Oscillation Workshop — PoS(NOW2022)

View full text Add to dashboard Cite

show abstract

“…These potential benefits have inspired research into proto-foundation models suitable for particle physics. For example, [24][25][26][27][28] investigated how known physical symmetries could be used to learn powerful embeddings of jet data, [29] showed the versatility of graph-based message passing networks for datasets from different domains of physics, [30,31] demonstrated conditioning generative models on the geometry of the detector to allow the simultaneous simulation of multiple detectors with one architecture, [32,33] used meta-learning for mass-decorrelation and weak-supervision, and [10] achieved state-of-the-art performance on the top tagging landscape dataset [34] by pre-training on a different dataset [35] and transferring the results.…”

Section: Introductionmentioning

confidence: 99%

OmniJet-α: the first cross-task foundation model for particle physics

Birk,

Hallin,

Kasieczka

2024

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

Foundation models are multi-dataset and multi-task machine learning methods that once pre-trained can be fine-tuned for a large variety of downstream applications. The successful development of such general-purpose models for physics data would be a major breakthrough as they could improve the achievable physics performance while at the same time drastically reduce the required amount of training time and data. We report significant progress on this challenge on several fronts. First, a comprehensive set of evaluation methods is introduced to judge the quality of an encoding from physics data into a representation suitable for the autoregressive generation of particle jets with transformer architectures (the common backbone of foundation models). These measures motivate the choice of a higher-fidelity tokenization compared to previous works. Finally, we demonstrate transfer learning between an unsupervised problem (jet generation) and a classic supervised task (jet tagging) with our new OmniJet-α model. This is the first successful transfer between two different and actively studied classes of tasks and constitutes a major step in the building of foundation models for particle physics.

show abstract

“…ML-methods for anomaly detection at the LHC have generally received a lot of attention in the context of anomalous jets [10][11][12][13][14][15][16][17], anomalous events pointing to physics beyond the Standard Model [18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35], or enhancing established search strategies [36][37][38][39][40][41][42]. They include a first ATLAS analysis [43], experimental validation of some of the methods [44,45], quantum machine learning [46], applications to heavy-ion collisions [47], the DarkMachines challenge [48], and the LHC Olympics 2020 community challenge [49,50].…”

Section: Introductionmentioning

confidence: 99%

What's anomalous in LHC jets?

Buss,

Dillon,

Finke

et al. 2023

SciPost Phys.

View full text Add to dashboard Cite

Searches for anomalies are a significant motivation for the LHC and help define key analysis steps, including triggers. We discuss specific examples how LHC anomalies can be defined through probability density estimates, evaluated in a physics space or in an appropriate neural network latent space, and discuss the model-dependence in choosing an appropriate data parameterisation. We illustrate this for classical k-means clustering, a Dirichlet variational autoencoder, and invertible neural networks. For two especially challenging scenarios of jets from a dark sector we evaluate the strengths and limitations of each method.

show abstract

Metalearning and data augmentation for mass-generalized jet taggers

Cited by 6 publications

References 40 publications

The T2K Near Detector Upgrade

The T2K Near Detector Upgrade

OmniJet-α: the first cross-task foundation model for particle physics

What's anomalous in LHC jets?

Contact Info

Product

Resources

About