One of the important aspects of searches for new physics at the Large Hadron Collider (LHC) involves the identification and reconstruction of single particles, jets and event topologies of interest in collision events. The End-to-End Deep Learning (E2E) project in the CMS experiment focuses on the development of these reconstruction and identification tasks with innovative deep learning approaches.
This project will focus on the development of foundation models for end-to-end particle reconstruction with the goal of performing generative, classification and regression tasks. Proposals can build on last year’s work which combined the L-GATr and ParT models and implemented a track-level masked autoencoder pretraining and present strategies to meaningfully improve on these techniques. This project is related to the project Event Classification With Masked Transformer Autoencoders but should focus primarily on developing complete multi-modal foundation model training pipelines.
Total project length: 175/350 hours
Advanced
C++, Python, PyTorch and some previous experience in Machine Learning.
Please use this link to access the test for this project.
Please DO NOT contact mentors directly by email. Questions should instead be directed to ml4-sci@cern.ch which is forwarded to mentors. To submit your proposal, CV, and test task solutions, please use this Google form.