Primordial correlators from multi-point propagators

Abstract

A key step in the comparison between inflationary predictions and cosmological observations is the computation of primordial correlators. Numerical methods have been developed that overcome some of the difficulties arising in analytical calculations when the models considered are complex. The PyTransport package, which implements the transport formalism, allows computation of the tree-level 2- and 3-point correlation functions for multi-field models with arbitrary potentials and a curved field space. In this work we investigate an alternative numerical implementation of the transport approach, based on the use of transfer “matrices” called multi-point propagators (MPP). We test the novel MPP method, and extensively compare it with the traditional implementation of the transport approach provided in PyTransport. We highlight advantages of the former, discussing its performance in terms of accuracy, precision and running time, as well as dependence on the number of e-folds of sub-horizon evolution and tolerance settings. For topical ultra-slow-roll models of inflation we show that MPPs (i) precisely track the decay of correlators even when PyTransport produces erroneous results, (ii) extend the computation of squeezed bispectra for squeezing values at least one decade beyond those attainable with PyTransport.