AstroAI Lunch Talks - February 9, 2026 - Ming-Shau Liu
09 Feb 2026 - Joshua Wing
The video can be found here: https://www.youtube.com/watch?v=_TAGC2-3P_4
Speaker: Ming-Shau Liu (Flatiron Institute)
Title: Continuous representations of baryonic feedback for robust inference
Abstract: Modelling baryonic feedback in cosmology is a nontrivial problem, and different simulation suites offer varied subgrid physics descriptions and parameterisations. We introduce a machine learning pipeline that captures a representation of the baryonic feedback, allowing for marginalisation over the theoretical uncertainties introduced by the simulation suites.