Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore how generative machine learning models are revolutionizing astrophysics research in this Institute for Advanced Study seminar by Carolina Cuesta-Lazaro. Discover how generative models enable high-dimensional simulation-based inference for cosmology, providing a roadmap to field-level inference through machine learning emulators that make computationally expensive hydrodynamical simulations accessible for large-scale analysis. Learn about applications to diverse cosmological observables including the kinetic and thermal Sunyaev-Zel'dovich effects, fast radio bursts, baryonification effects in weak lensing, and galaxy formation processes. Examine innovative methods for learning rich, low-dimensional representations that capture underlying physical processes in astrophysical data, including techniques that create joint latent spaces of shared and private information between simulations and observations. Understand approaches to disentangle instrumental systematics from underlying astrophysics in a fully data-driven manner, and explore the potential for LLM agents to close the loop in the scientific method by proposing, implementing, and testing cosmological theories.
