A sharp cut-off in the primordial scalar power spectrum on large scales has been known to improve the fit to the cosmic microwave background data when compared to the more standard, nearly scale invariant power spectrum that arises in slow roll inflation. Over the last couple of years, there has been a resurgent interest in arriving at such power spectra in models with kinetically dominated initial conditions for the background scalar field which leads to inflation of specific duration. In a recent work, we had numerically investigated the characteristics of the scalar bispectrum generated in such models. In this work, we compare the scenario with two other competing scenarios (viz. punctuated inflation and a model due to Starobinsky) which also suppress the scalar power in a roughly similar fashion on large scales. We further consider two other scenarios involving inflation of a finite duration, one wherein the scalar field begins on the inflationary attractor and another wherein the field starts with a smaller velocity and evolves towards the attractor. These scenarios too exhibit a sharp drop in power on large scales if the initial conditions on the perturbations for a range of modes are imposed on super-Hubble scales as in the kinetically dominated model. The model wherein the background field always remains on the inflationary attractor is interesting for the reason that it permits analytical calculations of the scalar power and bispectra. We compare the amplitudes and shapes of the scalar non-Gaussianity parameter f NL in all these cases which lead to scalar power spectra of similar form. Interestingly, we find that, in the models wherein the initial conditions on the perturbations are imposed on super-Hubble scales, the consistency relation governing the scalar bispectrum is violated for the large scale modes, whereas the relation is satisfied for all the modes in the other scenarios. These differences in the behavior of the scalar bispectra can conceivably help us observationally discriminate between the various models which lead to power spectra of roughly similar shape.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.