We demonstrate that light spectator fields can source sizeable CMB anisotropies through modulated reheating even in the absence of direct couplings to the inflaton. The effect arises when the phase space of the inflaton decay is modulated by the spectator which generates masses for the decay products. We call the mechanism \textit{indirect modulation} and show that it can source perturbations even four orders of magnitude larger than the observed. Importantly, the indirect mechanism is present in the Standard Model extended with right-handed neutrinos. For a minimally coupled Higgs boson this leads to a novel lower bound on the quartic coupling and constrains the neutrino Yukawas below unity.

We demonstrate that adiabatic dark matter can be generated by gravity induced symmetry breaking during inflation. We study a $Z_2$ symmetric scalar singlet that couples to other fields only through gravity and for which the symmetry is broken by the spacetime curvature during inflation when the non-minimal coupling $\xi$ is negative. We find that the symmetry breaking leads to the formation of adiabatic dark matter with the observed abundance for the singlet mass $m\sim{\rm MeV}$ and $|\xi|\sim 1$.

We demonstrate the existence of a generic, efficient and purely gravitational channel producing a significant abundance of dark relics during reheating after the end of inflation. The mechanism is present for any inert scalar with the non-minimal curvature coupling $\xi R\chi^2$ and the relic production is efficient for natural values $\xi = {\cal O}(1)$. The observed dark matter abundance can be reached for a broad range of relic masses extending from $m \sim 1 {\rm k eV}$ to $m \sim 10^{8} {\rm GeV}$, depending on the scale of inflation and the dark sector couplings. Frustratingly, such relics escape direct, indirect and collider searches since no non-gravitational couplings to visible matter are needed.

We show that the dynamics of the Higgs field during inflation is not affected by metric fluctuations if the Higgs is an energetically subdominant light spectator. For Standard Model parameters we find that couplings between Higgs and metric fluctuations are suppressed by $\mathcal{O}(10^{-7})$. They are negligible compared to both pure Higgs terms in the effective potential and the unavoidable non-minimal Higgs coupling to background scalar curvature. The question of the electroweak vacuum instability during high energy scale inflation can therefore be studied consistently using the Jordan frame action in a Friedmann--Lema\^itre--Robertson--Walker metric, where the Higgs-curvature coupling enters as an effective mass contribution. Similar results apply for other light spectator scalar fields during inflation.

We investigate the CMB $\mu$ distortion in models where two uncorrelated sources contribute to primordial perturbations. We parameterise each source by an amplitude, tilt, running and running of the running. We perform a detailed analysis of the distribution signal as function of the model parameters, highlighting the differences compared to single-source models. As a specific example, we also investigate the mixed inflaton-curvaton scenario. We find that the $\mu$ distortion could efficiently break degeneracies of curvaton parameters especially when combined with future sensitivity of probing the tensor-to-scalar ratio $r$. For example, assuming bounds $\mu < 0.5 \times 10^{-8}$ and $r<0.01$, the curvaton contribution should either vanish or the curvaton should dominate primordial perturbations and its slow-roll parameter $\eta_{\chi}$ is constrained to the interval $-0.007 < \eta_{\chi}< 0.045$.

We consider inflationary magnetogenesis where the conformal symmetry is broken by the term $f^2(\phi) F_{\alpha\beta} F^{\alpha\beta}$. We assume that the magnetic field power spectrum today between 0.1 and $10^4$ Mpc is a power law, with upper and lower limits from observation. This fixes $f$ to be close to a power law in conformal time in the window during inflation when the modes observed today are generated. In contrast to previous work, we do not make any assumptions about the form of $f$ outside these scales. We cover all possible reheating histories, described by an average equation of state $-1/3 <\bar{w} <1$. Requiring that strong coupling and large backreaction are avoided both at the background and perturbative level, we find the bound $\delta_{B_0} < 5 \times10^{-15} \left( \frac{r}{0.07} \right)^{1/2} \kappa \mathrm{G}$ for the magnetic field generated by inflation, where $r$ is the tensor-to-scalar ratio and $\kappa$ is a constant related to the form of $f$. This estimate has an uncertainty of one order of magnitude related to our approximations. The parameter $\kappa$ is $<100$, and values $\gtrsim1$ require a highly fine-tuned form of $f$; typical values are orders of magnitude smaller.

We show that curvature induced particle production at reheating generates adiabatic dark matter if there are non-minimally coupled spectator scalars weakly coupled to visible matter. The observed dark matter abundance implies an upper bound on spectator masses $m$ and non-minimal coupling values $\xi$. For example, assuming quadratic inflation, instant reheating and a single spectator scalar with only gravitational couplings, the observed dark matter abundance is obtained for $m\sim 0.1$ GeV and $\xi \sim 1$. Larger mass and coupling values of the spectator are excluded as they would lead to overproduction of dark matter.

We consider portal models which are ultraweakly coupled with the Standard Model, and confront them with observational constraints on dark matter abundance and isocurvature perturbations. We assume the hidden sector to contain a real singlet scalar $s$ and a sterile neutrino $\psi$ coupled to $s$ via a pseudoscalar Yukawa term. During inflation, a primordial condensate consisting of the singlet scalar $s$ is generated, and its contribution to the isocurvature perturbations is imprinted onto the dark matter abundance. We compute the total dark matter abundance including the contributions from condensate decay and nonthermal production from the Standard Model sector. We then use the Planck limit on isocurvature perturbations to derive a novel constraint connecting dark matter mass and the singlet self coupling with the scale of inflation: $m_{\rm DM}/{\rm GeV}\lesssim 0.2\lambda_{\rm s}^{\scriptscriptstyle 3/8} \left(H_*/10^{\scriptscriptstyle 11}{\rm GeV}\right)^{\scriptscriptstyle -3/2}$. This constraint is relevant in most portal models ultraweakly coupled with the Standard Model and containing light singlet scalar fields.

We investigate the dependency of Higgs inflation on the non-renormalisable matching between the low energy Standard Model limit and the inflationary regime at high energies. We show that for the top mass range $m_t \gtrsim 171.8$ GeV the scenario robustly predicts the spectral index $n_s \simeq 0.97$ and the tensor-to-scalar ratio $r\simeq 0.003$. The matching is however non-trivial, even the best-fit values $m_h=125.09$ GeV and $m_t=173.21$ GeV require a jump $\delta \lambda \sim 0.01$ in the Higgs coupling below the inflationary scale. For $m_t\lesssim 171.8$ GeV, the matching may generate a feature in the inflationary potential. In this case the predicted values of $n_s$ and $r$ vary but the model is still falsifiable. For example, a detection of negative running of spectral index at level $\alpha_s \lesssim -0.01$ would rule out Higgs inflation.

We study the resonant decay of the primordial Standard Model Higgs condensate after inflation into $SU(2)$ gauge bosons on the lattice. We find that the non-Abelian interactions between the gauge bosons quickly extend the momentum distribution towards high values, efficiently destroying the condensate after the onset of backreaction. For the inflationary scale $H = 10^8$ GeV, we find that 90% of the Higgs condensate has decayed after $n \sim 10$ oscillation cycles. This differs significantly from the Abelian case where, given the same coupling strengths, most of the condensate would persist after the resonance.

We investigate the dynamics of the Higgs field at the end of inflation in the minimal scenario consisting of an inflaton field coupled to the Standard Model only through the non-minimal gravitational coupling $\xi$ of the Higgs field. Such a coupling is required by renormalisation of the Standard Model in curved space, and in the current scenario also by vacuum stability during high-scale inflation. We find that for $\xi\gtrsim 1$, rapidly changing spacetime curvature at the end of inflation leads to significant production of Higgs particles, potentially triggering a transition to a negative-energy Planck scale vacuum state and causing an immediate collapse of the Universe.

We show that dark matter abundance and the inflationary scale $H$ could be intimately related. Standard Model extensions with Higgs mediated couplings to new physics typically contain extra scalars displaced from vacuum during inflation. If their coupling to Standard Model is weak, they will not thermalize and may easily constitute too much dark matter reminiscent to the moduli problem. As an example we consider Standard Model extended by a $Z_2$ symmetric singlet $s$ coupled to the Standard Model Higgs $\Phi$ via $\lambda \Phi^{\dag}\Phi s^2$. Dark matter relic density is generated non-thermally for $\lambda \lesssim 10^{-7}$. We show that the dark matter yield crucially depends on the inflationary scale. For $H\sim 10^{10}$ GeV we find that the singlet self-coupling and mass should lie in the regime $\lambda_{\rm s}\gtrsim 10^{-9}$ and $m_{\rm s}\lesssim 50$ GeV to avoid dark matter overproduction.

We discuss correlations among spectral observables as a new tool for differentiating between models for the primordial perturbation. We show that if generated in the isocurvature sector, a running of the scalar spectral index is correlated with the statistical properties of non-Gaussianities. In particular, we find a large running will inevitably be accompanied by a large running of $f_{\rm NL}$ and enhanced $g_{\rm NL}$, with $g_{\rm NL}\gg f_{\rm NL}^2$. If the tensor to scalar ratio is large, a large negative running must turn positive on smaller scales. Interestingly, the characteristic scale of the transition could potentially distinguish between the inflaton and isocurvature fields.

It has been claimed that the electroweak vacuum may be unstable during inflation due to large fluctuations of order $H$ in case of a high inflationary scale as suggested by BICEP2. We compute the Standard Model Higgs effective potential including UV-induced curvature corrections at one-loop level. We find that for a high inflationary scale a large curvature mass is generated due to RG running of non-minimal coupling $\xi$, which either stabilizes the potential against fluctuations for $\xi_{\rm EW} \gtrsim 6\cdot 10^{-2}$, or destabilizes it for $\xi_{\rm EW} \lesssim 2 \cdot 10^{-2}$ when the generated curvature mass is negative. Only in the narrow intermediate region the effect of the curvature mass may be significantly smaller.

We study the resonant decay of the Higgs condensate into weak gauge bosons after inflation and estimate the corrections arising from the non-Abelian self-interactions of the gauge fields. We find that non-Abelian interaction terms induce an effective mass which tends to shut down the resonance. For the broad resonance relevant for the Standard Model Higgs the produced gauge particles backreact on the dynamics of the Higgs condensate before the non-Abelian terms grow large. The non-Abelian terms can however significantly affect the final stages of the resonance after the backreaction. In the narrow resonance regime, which may be important for extensions of the Standard Model, the non-Abelian terms affect already the linear stage and terminate the resonance before the Higgs condensate is affected by the backreaction of decay products.

We study the post-inflationary dynamics of the Standard Model Higgs and a real singlet scalar $s$, coupled together through a renormalizable coupling $\lambda_{sh}h^2s^2$, in a $Z_2$ symmetric model that may explain the observed dark matter abundance and/or the origin of baryon asymmetry. The initial values for the Higgs and $s$ condensates are given by inflationary fluctuations, and we follow their dissipation and relaxation to the low energy vacua. We find that both the lowest order perturbative and the non-perturbative decays are blocked by thermal effects and large background fields and that the condensates decay by two-loop thermal effects. Assuming instant reheating at $T=10^{16}$ GeV, the characteristic temperature for the Higgs condensate thermalization is found to be $T_h \sim 10^{14}$ GeV, whereas $s$ thermalizes typically around $T_s \sim 10^{6}$ GeV. By that time, the amplitude of the singlet is driven very close to the vacuum value by the expansion of the universe, unless the portal coupling takes a value $\lambda_{sh}\lesssim 10^{-7}$ and the singlet $s$ never thermalizes. With these values of the coupling, it is possible to slowly produce a sizeable fraction of the observed dark matter abundance via singlet condensate fragmentation and thermal Higgs scattering. Physics also below the electroweak scale can therefore be affected by the non-vacuum initial conditions generated by inflation.

We investigate inflationary Higgs dynamics and constraints on the Standard Model parameters assuming the Higgs potential, computed to next-to-next leading order precision, is not significantly affected by new physics. For a high inflationary scale $H\sim 10^{14}$ GeV suggested by BICEP2, we show that the Higgs is a light field subject to fluctuations which affect its dynamics in a stochastic way. Starting from its inflationary value the Higgs must be able to relax to the Standard Model vacuum well before the electroweak scale. We find that this is consistent with the high inflationary scale only if the top mass $m_t$ is significantly below the best fit value. The region within $2\sigma$ errors of the measured $m_t$, the Higgs mass $m_h$ and the strong coupling $\alpha_s$ and consistent with inflation covers approximately the interval $m_t \lesssim 171.8\,{\rm GeV} + 0.538(m_h-125.5\,{\rm GeV})$ with $125.4\,{\rm GeV}\lesssim m_h\lesssim 126.3\,{\rm GeV}$. If the low top mass region could be definitively ruled out, the observed high inflationary scale alone, if confirmed, would seem to imply new physics necessarily modifying the Standard Model Higgs potential below the inflationary scale.

In order to gain new insights into the gauge field couplings in the early universe, we consider the constraints on gauge field production during inflation imposed by requiring that their effect on the CMB anisotropies are subdominant. In particular, we calculate systematically the bispectrum of the primordial curvature perturbation induced by the presence of vector gauge fields during inflation. Using a model independent parametrization in terms of magnetic non-linearity parameters, we calculate for the first time the contribution to the bispectrum from the cross correlation between the inflaton and the magnetic field defined by the gauge field. We then demonstrate that in a very general class of models, the bispectrum induced by the cross correlation between the inflaton and the magnetic field can be dominating compared with the non-Gaussianity induced by magnetic fields when the cross correlation between the magnetic field and the inflaton is ignored.

We investigate the dynamics of the Standard Model higgs with a minimal coupling to gravity during and after inflation. In the regime where the Standard Model vacuum is stable, we find that the higgs becomes a light spectator field after about 30 efolds of inflation, irrespectively of its initial value. Once the higgs has become light, its root-mean-square value h_* relaxes to equilibrium in about 85 efolds for the inflationary scale of H_*=10^4 GeV and in 20 efolds for H_*=10^10 GeV. The equilibrium value is given by h_*~0.36 lambda^{-1/4}H_*, where lambda=0.09 ... 0.0005 is the higgs self coupling at the scales H_*=10^4 ... 10^10 GeV. We show that the main decay channel of the higgs condensate after inflation is the resonant production of Standard Model gauge bosons. For a set of parameters we find that a significant part of the condensate has decayed in between 340 and 630 Hubble times after the onset of higgs oscillations, depending on H_* in a non-trivial way. The higgs perturbations correspond to isocurvature modes during inflation but they could generate significant adiabatic perturbations at a later stage for example through a modulation of the reheating stage. However, this requires that the inflaton(s) decay no later than a few hundred Hubble times after the onset of higgs oscillations.

The new Planck constraints on the local bispectrum parameter fnl are about 10^5 times tighter than the current constraints on the trispectrum parameter gnl, which means that the allowed numerical values of the second and third order terms in the perturbative expansion of the curvature perturbation are comparable. We show that a consequence of this is that if gnl is large enough to be detectable, then it will induce a large variation between the observable value of fnl and its value in a larger inflated volume. Even if there were only a few extra efoldings between the beginning of inflation and horizon crossing of our Hubble horizon, an observably large gnl means that fnl is unlikely to be as small as its current constraint, regardless of its true background value. This result is very general, it holds regardless of how many fields contributed to the curvature perturbation. We also generalise this result to other shapes of non-Gaussianity, beyond the local model. We show that the variance of the 3-point function in the squeezed limit is bounded from below by the square of the squeezed limit of the 4-point function.

Inflation is the leading paradigm for explaining the origin of primordial density perturbations and the observed temperature fluctuations of the cosmic microwave background. However many open questions remain, in particular whether one or more scalar fields were present during inflation and how they contributed to the primordial density perturbation. We propose a new observational test of whether multiple fields, or only one (not necessarily the inflaton) generated the perturbations. We show that our test, relating the bispectrum and trispectrum, is protected against loop corrections at all orders, unlike previous relations.

Primordial perturbations with wavelengths greater than the observable universe shift the effective background fields in our observable patch from their global averages over the inflating space. This leads to a landscape picture where the properties of our observable patch depend on its location and may significantly differ from the expectation values predicted by the underlying fundamental inflationary model. We show that if multiple fields are present during inflation, this may happen even if our horizon exit would be preceded by only a few e-foldings of inflation. Non-Gaussian statistics are especially affected: for example models of local non-Gaussianity predicting |f_NL|>> 10 over the entire inflating volume can have a probability up to a few tens of percent to generate a non-detectable bispectrum in our observable patch |fNL^{obs.}|<10. In this work we establish systematic connections between the observable local properties of primordial perturbations and the global properties of the inflating space which reflect the underlying high energy physics. We study in detail the implications of both a detection and non-detection of primordial non-Gaussianity by Planck, and discover novel ways of characterising the naturalness of different observational configurations.

We propose a method to probe higher-order correlators of the primordial density field through the inhomogeneity of local non-Gaussian parameters, such as f_NL, measured within smaller patches of the sky. Correlators between n-point functions measured in one patch of the sky and k-point functions measured in another patch depend upon the (n+k)-point functions over the entire sky. The inhomogeneity of non-Gaussian parameters may be a feasible way to detect or constrain higher-order correlators in local models of non-Gaussianity, as well as to distinguish between single and multiple-source scenarios for generating the primordial density perturbation, and more generally to probe the details of inflationary physics.

Cosmological runaway solutions may exhibit an exact dilatation symmetry in the asymptotic limit of infinite time. In this limit, the massless dilaton or cosmon could be accompanied by another massless scalar field - the geon. At finite time, small time-dependent masses for both the cosmon and geon are still present due to imperfect dilatation symmetry. For a sufficiently large mass the geon will start oscillating and play the role of dark matter, while the cosmon is responsible for dark energy. The common origin of the mass of both fields leads to an effective interaction between dark matter and dark energy. Realistic cosmologies are possible for a simple form of the effective cosmon-geon-potential. We find an inverse geon mass of a size where it could reduce subgalactic structure formation.

We study the scale dependence of the non-linearity parameters f_NL and g_NL in curvaton models with self-interactions. We show that the spectral indices n_fNL=d ln|f_NL|/(d ln k) and n_gNL=d ln |g_NL|/(d ln k) can take values much greater than the slow--roll parameters and the spectral index of the power spectrum. This means that the scale--dependence of the bi and trispectrum could be easily observable in this scenario with Planck, which would lead to tight additional constraints on the model. Inspite of the highly non-trivial behaviour of f_NL and g_NL in the curvaton models with self-interactions, we find that the model can be falsified if g_NL(k) is also observed.

We calculate the scale dependence of the bispectrum and trispectrum in (quasi) local models of non-Gaussian primordial density perturbations, and characterize this scale dependence in terms of new observable parameters. They can help to discriminate between models of inflation, since they are sensitive to properties of the inflationary physics that are not probed by the standard observables. We find consistency relations between these parameters in certain classes of models. We apply our results to a scenario of modulated reheating, showing that the scale dependence of non-Gaussianity can be significant. We also discuss the scale dependence of the bispectrum and trispectrum, in cases where one varies the shape as well as the overall scale of the figure under consideration. We conclude providing a formulation of the curvature perturbation in real space, which generalises the standard local form by dropping the assumption that f_NL and g_NL are constants.

We describe a simple way of incorporating fluctuations of the Hubble scale during the horizon exit of scalar perturbations into the delta N formalism. The dominant effect comes from the dependence of the Hubble scale on low-frequency modes of the inflaton. This modifies the coefficient of the log-enhanced term appearing in the curvature spectrum at second order in field fluctuations. With this modification, the relevant coefficient turns out to be proportional to the second derivative of the tree-level spectrum with respect to the inflaton phi at horizon exit. A logarithm with precisely the same coefficient appears in a calculation of the log-enhancement of the curvature spectrum based purely on the geometry of the reheating surface. We take this agreement as strong support for the proposed implementation of the delta N formalism. Moreover, our analysis makes it apparent that the log-enhancement of the inflationary power-spectrum is indeed physical if this quantity is defined using a global coordinate system on the reheating surface (or any other post-inflationary surface of constant energy density). However, it can be avoided by defining the spectrum using invariant distances on this surface.

We calculate curvature perturbations in the scenario in which the curvaton field decays into another scalar field via parametric resonance. As a result of a nonlinear stage at the end of the resonance, standard perturbative calculation techniques fail in this case. Instead, we use lattice field theory simulations and the separate universe approximation to calculate the curvature perturbation as a nonlinear function of the curvaton field. For the parameters tested, the generated perturbations are highly non-Gaussian and not well approximated by the usual fNL parameterisation. Resonant decay plays an important role in the curvaton scenario and can have a substantial effect on the resulting perturbations.

We consider possible scale-dependence of the non-linearity parameter f_NL in local and quasi-local models of non-Gaussian primordial density perturbations. In the simplest model where the primordial perturbations are a quadratic local function of a single Gaussian field then f_NL is scale-independent by construction. However scale-dependence can arise due to either a local function of more than one Gaussian field, or due to non-linear evolution of modes after horizon-exit during inflation. We show that the scale dependence of f_NL is typically first order in slow-roll. For some models this may be observable with experiments such as Planck provided that f_NL is close to the current observational bounds.

We present a systematic study of the amplitude of the primordial perturbation in curvaton models with self-interactions, treating both renormalizable and non-renormalizable interactions. In particular, we consider the possibility that the curvaton energy density is subdominant at the time of the curvaton decay. We find that large regions in the parameter space give rise to the observed amplitude of primordial perturbation even for non-renormalizable curvaton potentials, for which the curvaton energy density dilutes fast. At the time of its decay, the curvaton energy density may typically be subdominant by a relative factor of 10^-3 and still produce the observed perturbation. Field dynamics turns out to be highly non-trivial, and for non-renormalizable potentials and certain regions of the parameter space we observe a non-monotonous relation between the final curvature perturbation and the initial curvaton value. In those cases, the time evolution of the primordial perturbation also displays an oscillatory behaviour before the curvaton decay.

We investigate non-Gaussianities in self-interacting curvaton models treating both renormalizable and non-renormalizable polynomial interactions. We scan the parameter space systematically and compute numerically the non-linearity parameters f_NL and g_NL. We find that even in the interaction dominated regime there are large regions consistent with current observable bounds. Whenever the interactions dominate, we discover significant deviations from the relations f_NL ~ 1/r_decay and g_NL ~ 1/r_decay valid for quadratic curvaton potentials, where r_decay measures the curvaton contribution to the total energy density at the time of its decay. Even if r_decay << 1, there always exists regions with f_NL ~ 0 since the sign of f_NL oscillates as a function of the parameters. While g_NL can also change sign, typically g_NL is non-zero in the low-f_NL regions. Hence, for some parameters the non-Gaussian statistics is dominated by g_NL rather than by f_NL. Due to self-interactions, both the relative signs of f_NL and g_NL and the functional relation between them is typically modified from the quadratic case, offering a possible experimental test of the curvaton interactions.

We argue that the curvaton decay takes place most naturally by way of a broad parametric resonance. The mechanism is analogous to resonant inflaton decay but does not require any tuning of the curvaton coupling strength to other scalar fields. For low scale inflation and a correspondingly low mass scale for the curvaton, we speculate on observable consequences including the possibility of stochastic gravitational waves

We discuss the infrared divergences that appear to plague cosmological perturbation theory. We show that within the stochastic framework they are regulated by eternal inflation so that the theory predicts finite fluctuations. Using the $\Delta N$ formalism to one loop, we demonstrate that the infrared modes can be absorbed into additive constants and the coefficients of the diagrammatic expansion for the connected parts of two and three-point functions of the curvature perturbation. As a result, the use of any infrared cutoff below the scale of eternal inflation is permitted, provided that the background fields are appropriately redefined. The natural choice for the infrared cutoff would of course be the present horizon; other choices manifest themselves in the running of the correlators. We also demonstrate that it is possible to define observables that are renormalization group invariant. As an example, we derive a non-perturbative, infrared finite and renormalization point independent relation between the two-point correlators of the curvature perturbation for the case of the free single field.

We consider the supergravity origin of the recently proposed MSSM inflationary model, which relies on the existence of a saddle point along a dimension six flat direction. We derive the conditions that the Kahler potential has to satisfy for the saddle point to exist irrespective of the hidden sector vevs. We show that these conditions are satisfied by a simple class of Kahler potentials, which we find to have a similar form as in various string theory compactifications. For these potentials, slow roll MSSM inflation requires no fine tuning of the soft supersymmetry breaking parameters.

Recently it has been argued that some of the fine-tuning problems of the MSSM inflation associated with the existence of a saddle point along a flat direction may be solved naturally in a class of supergravity models. Here we extend the analysis and show that the constraints on the Kahler potentials in these models are considerably relaxed when the location of the saddle point is treated as a free variable. We also examine the effect of supergravity corrections on inflationary predictions and find that they can slightly alter the value of the spectral index. As an example, for flat direction field values $|\bar{\phi}_0|=1\times10^{-4}M_P$ we find $n\sim0.92 ... 0.94$ while the prediction of the MSSM inflation without any corrections is $n\sim0.92$.

We present an approach to cosmological perturbations based on a covariant perturbative expansion between two worldlines in the real inhomogeneous universe. As an application, at an arbitrary order we define an exact scalar quantity which describes the inhomogeneities in the number of e-folds on uniform density hypersurfaces and which is conserved on all scales for a barotropic ideal fluid. We derive a compact form for its conservation equation at all orders and assign it a simple physical interpretation. To make a comparison with the standard perturbation theory, we develop a method to construct gauge-invariant quantities in a coordinate system at arbitrary order, which we apply to derive the form of the n-th order perturbation in the number of e-folds on uniform density hypersurfaces and its exact evolution equation. On large scales, this provides the gauge-invariant expression for the curvature perturbation on uniform density hypersurfaces and its evolution equation at any order.

We consider curvaton models with potentials that depart slightly from the quadratic form. We show that although such a small departure does not modify significantly the Gaussian part of the curvature perturbation, it can have a pronounced effect on the level of non-Gaussianity. We find that unlike in the quadratic case, the limit of small non-Gaussianity, $|f_{NL}|\ll1$, is quite possible even with small curvaton energy density $r\ll1$ . Furthermore, non-Gaussianity does not imply any strict bounds on $r$ but the bounds depend on the assumptions about the higher order terms in the curvaton potential.