galaxies Black Holes Developing subgrid models for BH formation, motion, accretion, and feedback Resolved Star Formation and Galactic Winds Developing subgrid models for the star-forming interstellar medium and galactic winds Synthetic Observations Creating realistic galaxy and CMB observables forward models Accelerated Forward Modeling Using deep learning to create rapid emulators and inference tools Cosmology Developing cosmological techniques with advanced subgrid techniques to carry out large cosmological simulations for training and analysis. Training Set Generation Designing large suites of cosmological simulations inference Explicit Likelihood (BORG) Inferring cosmological initial conditions Implicit Likelihood Inference Developing techniques to infer information from astronomical observations Robustness / Interpretability Studying model misspecification robustness and model interpretability