Accelerating New Physics Discovery with Foundation Models

Funder	Science and Technology Facilities Council
Recipient Organization	University of Cambridge
Country	United Kingdom
Start Date	Sep 30, 2024
End Date	Mar 30, 2028
Duration	1,277 days
Number of Grantees	1
Roles	Student
Data Source	UKRI Gateway to Research
Grant ID	`2927200`

Grant Description

This thesis will look to develop and apply innovative machine learning techniques for discovering new models for describing physical phenomena across the physical sciences, from cosmology to magnetohydrodynamics.

This thesis will take place embedded within the international 'PolymathicAI' collaboration which seeks to build a foundation model for science, with the goal of disentangling new mathematical frameworks that cut across disciplines in novel ways.

These methods will be applied to specific problems in cosmology and MHD simulations, with the aim of discovering new physical insights.