{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/energy-flow-networks-deep-sets-for-particle","title":"Energy Flow Networks: Deep Sets for Particle Jets","arxiv_id":"1810.05165","date":"2018-10-11","proceeding":null,"authors":["Patrick T. Komiske","Eric M. Metodiev","Jesse Thaler"],"abstract":"A key question for machine learning approaches in particle physics is how to\nbest represent and learn from collider events. As an event is intrinsically a\nvariable-length unordered set of particles, we build upon recent machine\nlearning efforts to learn directly from sets of features or \"point clouds\".\nAdapting and specializing the \"Deep Sets\" framework to particle physics, we\nintroduce Energy Flow Networks, which respect infrared and collinear safety by\nconstruction. We also develop Particle Flow Networks, which allow for general\nenergy dependence and the inclusion of additional particle-level information\nsuch as charge and flavor. These networks feature a per-particle internal\n(latent) representation, and summing over all particles yields an overall\nevent-level latent representation. We show how this latent space decomposition\nunifies existing event representations based on detector images and radiation\nmoments. To demonstrate the power and simplicity of this set-based approach, we\napply these networks to the collider task of discriminating quark jets from\ngluon jets, finding similar or improved performance compared to existing\nmethods. We also show how the learned event representation can be directly\nvisualized, providing insight into the inner workings of the model. These\narchitectures lend themselves to efficiently processing and analyzing events\nfor a wide variety of tasks at the Large Hadron Collider. Implementations and\nexamples of our architectures are available online in our EnergyFlow package.","url_abs":"http://arxiv.org/abs/1810.05165v2","url_pdf":"http://arxiv.org/pdf/1810.05165v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"energy-flow-networks-deep-sets-for-particle","repo_url":"https://github.com/jet-universe/particle_transformer","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"energy-flow-networks-deep-sets-for-particle","repo_url":"https://github.com/womogenes/photon-jet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"machine-learning","task_name":"BIG-bench Machine Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1810.05165","atlas_url":"https://app.syntology.ai/?focus=1810.05165","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}