Facets of Gravity: A journey exploring spacetime from Planckian to cosmological scales

Abstract

What does it mean to quantize gravity, and what can a quantum theory of spacetime actually predict? This thesis explores these questions through complementary non-perturbative approaches, from the geometry of quantum spacetime to the emergence of cosmological physics. Within Causal Dynamical Triangulations, we develop observables to investigate effective topology, statistical homogeneity, and curvature in strongly fluctuating geometries, where conventional differential-geometric tools are unavailable. Tests in two-dimensional models reveal distinct coarse-graining behaviours and provide new ways to probe the emergence of familiar geometric properties. We then turn to the Functional Renormalization Group and Asymptotic Safety, developing flow invariants to characterize ultraviolet-critical surfaces and introducing Flow Invariant Neural Networks as a computational method for reconstructing them. In scalar-tensor gravity, we investigate how ultraviolet consistency conditions can constrain infrared interactions, obtaining an inflationary scenario compatible with cosmological observations. Finally, we question whether the metric must be the fundamental variable of quantum gravity. We formulate a lattice statistical approach to purely affine gravity, in which the connection is fundamental and the metric, when it exists, is emergent. Monte Carlo simulations provide exploratory results in two dimensions, while a four-dimensional homogeneous and isotropic construction yields preliminary evidence for de Sitter-like geometry and allows the study of curvature correlations. Together, these investigations develop complementary tools for studying quantum spacetime, testing its possible continuum behaviour, and connecting fundamental gravitational theories with observable physics, while identifying the conceptual and computational challenges that remain.