In this thesis we will primarily demonstrate how additional scalar degrees of freedom - which are motivated from many high-energy embeddings - open up new observational windows onto the physics of inflation. We construct a Bayesian framework to statistically compare models with additional fields given the current astronomical data. Putting inflation to the test, we perform our analysis on the quadratic curvaton accompanying a range of inflationary potentials, where we find that only one potential remains as a viable candidate. Furthermore, if the curvaton mechanism were to be confirmed by future non-Gaussianity measurements (from large scale structure surveys), the model could prove to be tremendously informative of the early inflationary history. The initial conditions given to these scalar fields become apparent when considering their fundamentally quantum behaviour. Taking this physics into account leads us to develop detailed models for post-inflationary phenomenology (namely, the curvaton and freeze-in dark matter models) and to discover powerful new probes of inflation itself. We further demonstrate how this theoretical study complements our statistical approach by motivating the prior information in our Bayesian analyses. The thesis finishes with a discussion of the future prospects for inflationary model selection. By hypothesising different toy survey configurations, we forecast different outcomes using information theory and our newly developed Bayesian experimental design formalism. In particular, we find that the most likely observable to optimise model selection between single-field inflationary models, through an order of magnitude precision improvement in the future, is the scalar spectral index. We conclude with a summary of the results obtained throughout.

Simple models of single-field inflation in the very early universe can generate the observed amplitude and scale dependence of the primordial density perturbation, but models with multiple fields can provide an equally good fit to current data. We show how future observations will be able to distinguish between currently favoured models. If a curvaton field generates the primordial perturbations after inflation, we show how the total duration of inflation can be measured.

We study the scalar-tensor-tensor non-Gaussian signal in an inflationary model comprising also an axion coupled with SU(2) gauge fields. In this set-up, metric fluctuations are sourced by the gauge fields already at the linear level providing an enhanced chiral gravitational waves spectrum. The same mechanism is at work in generating an amplitude for the three-point function that is parametrically larger than in standard single-field inflation.

How much more will we learn about single-field inflationary models in the future? We address this question in the context of Bayesian design and information theory. We develop a novel method to compute the expected utility of deciding between models and apply it to a set of futuristic measurements. This necessarily requires one to evaluate the Bayesian evidence many thousands of times over, which is numerically challenging. We show how this can be done using a number of simplifying assumptions and discuss their validity. We also modify the form of the expected utility, as previously introduced in the literature in different contexts, in order to partition each possible future into either the rejection of models at the level of the maximum likelihood or the decision between models using Bayesian model comparison. We then quantify the ability of future experiments to constrain the reheating temperature and the scalar running. Our approach allows us to discuss possible strategies for maximising information from future cosmological surveys. In particular, our conclusions suggest that, in the context of inflationary model selection, a decrease in the measurement uncertainty of the scalar spectral index would be more decisive than a decrease in the uncertainty in the tensor-to-scalar ratio. We have incorporated our approach into a publicly available python class, foxi (https://sites.google.com/view/foxicode), that can be readily applied to any survey optimisation problem.

We show that in the Feebly Interacting Massive Particle (FIMP) model of Dark Matter (DM), one may express the inflationary energy scale $H_*$ as a function of three otherwise unrelated quantities, the DM isocurvature perturbation amplitude, its mass and its self-coupling constant, independently of the tensor-to-scalar ratio. The FIMP model assumes that there exists a real scalar particle that alone constitutes the DM content of the Universe and couples to the Standard Model via a Higgs portal. We consider carefully the various astrophysical, cosmological and model constraints, accounting also for variations in inflationary dynamics and the reheating history, to derive a robust estimate for $H_*$ that is confined to a relatively narrow range. We point out that, within the context of the FIMP DM model, one may thus determine $H_*$ reliably even in the absence of observable tensor perturbations.

We study the stochastic distribution of spectator fields predicted in different slow-roll inflation backgrounds. Spectator fields have a negligible energy density during inflation but may play an important dynamical role later, even giving rise to primordial density perturbations within our observational horizon today. During de-Sitter expansion there is an equilibrium solution for the spectator field which is often used to estimate the stochastic distribution during slow-roll inflation. However slow roll only requires that the Hubble rate varies slowly compared to the Hubble time, while the time taken for the stochastic distribution to evolve to the de-Sitter equilibrium solution can be much longer than a Hubble time. We study both chaotic (monomial) and plateau inflaton potentials, with quadratic, quartic and axionic spectator fields. We give an adiabaticity condition for the spectator field distribution to relax to the de-Sitter equilibrium, and find that the de-Sitter approximation is never a reliable estimate for the typical distribution at the end of inflation for a quadratic spectator during monomial inflation. The existence of an adiabatic regime at early times can erase the dependence on initial conditions of the final distribution of field values. In these cases, spectator fields acquire sub-Planckian expectation values. Otherwise spectator fields may acquire much larger field displacements than suggested by the de-Sitter equilibrium solution. We quantify the information about initial conditions that can be obtained from the final field distribution. Our results may have important consequences for the viability of spectator models for the origin of structure, such as the simplest curvaton models.

Inflation in the early Universe is one of the most promising probes of gravity in the high-energy regime. However, observable scales give access to a limited window in the inflationary dynamics. In this essay, we argue that quantum corrections to the classical dynamics of cosmological fields allow us to probe much earlier epochs of the inflationary phase and extend this window by many orders of magnitude. We point out that both the statistics of cosmological fluctuations at observable scales, and the field displacements acquired by spectator fields that play an important role in many post-inflationary processes, are sensitive to a much longer phase of the inflationary epoch.

We derive the first systematic observational constraints on reheating in models of inflation where an additional light scalar field contributes to primordial density perturbations and affects the expansion history during reheating. This encompasses the original curvaton model but also covers a larger class of scenarios. We find that, compared to the single-field case, lower values of the energy density at the end of inflation and of the reheating temperature are preferred when an additional scalar field is introduced. For instance, if inflation is driven by a quartic potential, which is one of the most favoured models when a light scalar field is added, the upper bound $T_{\mathrm{reh}}<5\times 10^{4}\,\mathrm{GeV}$ on the reheating temperature $T_{\mathrm{reh}}$ is derived, and the implications of this value on post-inflationary physics are discussed. The information gained about reheating is also quantified and it is found that it remains modest in plateau inflation (though still larger than in the single-field version of the model) but can become substantial in quartic inflation. The role played by the vev of the additional scalar field at the end of inflation is highlighted, and opens interesting possibilities for exploring stochastic inflation effects that could determine its distribution.

We perform a Bayesian model comparison for scenarios within the quadratic curvaton model, determining the degree to which both are disfavoured with respect to the $\Lambda$CDM concordance model and single-field quadratic inflation, using the recent \emph{Planck} data release. Despite having three additional model parameters, the simplest curvaton scenario is not disfavoured relative to single-field quadratic inflation, and it becomes favoured against this single-field model when we include the joint BICEP/Keck/\emph{Planck} analysis. In all cases we assume an instantaneous inflaton decay and no surviving isocurvature perturbations. Despite the success of \emph{Planck} reaching its forecast measurement accuracy, we show that the current constraints on local non-Gaussianity are insufficiently precise to have any significant impact on the evidence ratios so far. We also determine the precision $\sigma(f_{\mathrm{NL}})$ required by future measurements assuming a fiducial value of $f_{\mathrm{NL}}=-5/4$ or $10.8$ to no longer disfavour the curvaton against the $\Lambda$CDM parametrisation, and we discuss the effect that the predicted increase in precision from future measurements on $f_{\mathrm{NL}}$ may have. We show that our results are not very sensitive to our choice of priors.