Constant roll and non-Gaussian tail in light of logarithmic duality
Ryoto Inui, Hayato Motohashi, Shi Pi, Yuichiro Tada, Shuichiro, Yokoyama

TL;DR
This paper explores how constant-roll inflation models exhibit non-Gaussian tails in curvature perturbations, linked to a logarithmic duality, impacting primordial black hole formation estimates.
Contribution
It reveals the role of logarithmic duality in constant-roll inflation and characterizes the non-Gaussian tail behavior of curvature perturbations.
Findings
Critical value $eta=-3/2$ determines stability of the CR condition.
Curvature perturbations follow a logarithmic mapping of Gaussian fields.
Tail behaviors include exponential and Gumbel-like distributions.
Abstract
The curvature perturbation in a model of constant-roll (CR) inflation is interpreted in view of the logarithmic duality discovered in Ref. [1] according to the formalism. We confirm that the critical value determining whether the CR condition is stable or not is understood as the point at which the dual solutions, i.e., the attractor and non-attractor solutions of the field equation, are interchanged. For the attractor-solution domination, the curvature perturbation in the CR model is given by a simple logarithmic mapping of a Gaussian random field, which can realise both the exponential tail (i.e., the single exponential decay) and the Gumbel-distribution-like tail (i.e., the double exponential decay) of the probability density function, depending on the value of . Such a tail behaviour is important for, e.g., the…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsStochastic processes and financial applications
