[Cosmology seminar] Aletheia: unlocking the non-linear Universe with evolution mapping

The standard model of cosmology, LCDM, is facing its most significant challenge yet. Results from the Dark Energy Spectroscopic Instrument (DESI) present tantalising evidence for dynamic dark energy. However, our ability to confirm such deviations is severely limited by a critical bottleneck: our current failure to accurately model the information-rich, non-linear cosmic web. Current analysis frameworks break down on non-linear scales, forcing us to discard vast amounts of data from flagship surveys. Evolution Mapping offers a path forward by reframing non-linear evolution by separating cosmological parameters into those governing the initial shape of the linear power spectrum and those controlling its subsequent growth. By compressing all growth parameters into a single "master clock”, this framework drastically simplifies the challenge of modelling the non-linear Universe. In this talk, I will introduce Aletheia, a next-generation emulator built on this principle that achieves 0.2% accuracy in predicting the matter power spectrum, P(k), representing a five-fold improvement over previous state-of-the-art models. I will illustrate how the same principle can be applied to model other non-linear statistics of the large-scale structure and integrated into deep learning architectures for Simulation-Based Inference (SBI), providing the essential toolkit for the upcoming legacy data harvest of Euclid and DESI. Finally, I will discuss how the evolution mapping framework provides a clearer understanding of the physical information content of cosmological observables.