Exploring the Intersection of Physics Modeling and Representation Learning
Representation Learning has evolved into a multi-purpose tool capable of solving arbitrary problems provided enough data. This thesis focuses on two primary directions: (1) Harnessing the power of deep learning for applications in fundamental physics and (2) using physicsinspired tools to improve an...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Published: |
Massachusetts Institute of Technology
2024
|
| Online Access: | https://hdl.handle.net/1721.1/157597 |
| _version_ | 1824458433196195840 |
|---|---|
| author | Kitouni, Ouail |
| author2 | Williams, Mike |
| author_facet | Williams, Mike Kitouni, Ouail |
| author_sort | Kitouni, Ouail |
| collection | MIT |
| description | Representation Learning has evolved into a multi-purpose tool capable of solving arbitrary problems provided enough data. This thesis focuses on two primary directions: (1) Harnessing the power of deep learning for applications in fundamental physics and (2) using physicsinspired tools to improve and shed some light on otherwise large-scale, inscrutable black-box algorithms. We explore a collection of applications that improve different aspects of nuclear and particle physics research across its many stages, from online data selection to offline data analysis. We also tease out how deep learning can open up entirely new avenues of research through the lens of mechanistic interpretability to (re)derive fundamental theory as well as new ways to reinterpret physics measurements. Lastly, we study how physics tools can be useful to better understand the dynamics of deep learning as well as provide a solid foundation for algorithms and training paradigms that expand the frontier of machine learning. |
| first_indexed | 2025-02-19T04:25:49Z |
| format | Thesis |
| id | mit-1721.1/157597 |
| institution | Massachusetts Institute of Technology |
| last_indexed | 2025-02-19T04:25:49Z |
| publishDate | 2024 |
| publisher | Massachusetts Institute of Technology |
| record_format | dspace |
| spelling | mit-1721.1/1575972024-11-19T03:03:25Z Exploring the Intersection of Physics Modeling and Representation Learning Kitouni, Ouail Williams, Mike Massachusetts Institute of Technology. Department of Physics Representation Learning has evolved into a multi-purpose tool capable of solving arbitrary problems provided enough data. This thesis focuses on two primary directions: (1) Harnessing the power of deep learning for applications in fundamental physics and (2) using physicsinspired tools to improve and shed some light on otherwise large-scale, inscrutable black-box algorithms. We explore a collection of applications that improve different aspects of nuclear and particle physics research across its many stages, from online data selection to offline data analysis. We also tease out how deep learning can open up entirely new avenues of research through the lens of mechanistic interpretability to (re)derive fundamental theory as well as new ways to reinterpret physics measurements. Lastly, we study how physics tools can be useful to better understand the dynamics of deep learning as well as provide a solid foundation for algorithms and training paradigms that expand the frontier of machine learning. Ph.D. 2024-11-18T19:13:26Z 2024-11-18T19:13:26Z 2024-09 2024-10-24T16:13:15.770Z Thesis https://hdl.handle.net/1721.1/157597 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) Copyright retained by author(s) https://creativecommons.org/licenses/by-nc-nd/4.0/ application/pdf Massachusetts Institute of Technology |
| spellingShingle | Kitouni, Ouail Exploring the Intersection of Physics Modeling and Representation Learning |
| title | Exploring the Intersection of Physics Modeling and Representation Learning |
| title_full | Exploring the Intersection of Physics Modeling and Representation Learning |
| title_fullStr | Exploring the Intersection of Physics Modeling and Representation Learning |
| title_full_unstemmed | Exploring the Intersection of Physics Modeling and Representation Learning |
| title_short | Exploring the Intersection of Physics Modeling and Representation Learning |
| title_sort | exploring the intersection of physics modeling and representation learning |
| url | https://hdl.handle.net/1721.1/157597 |
| work_keys_str_mv | AT kitouniouail exploringtheintersectionofphysicsmodelingandrepresentationlearning |