\documentclass[12pt]{article}
\usepackage{amsmath,amssymb,amsfonts}
\begin{document}
The BICEP2 experiment [arXiv:1403.3985] has announced a signal for primordial gravity waves with tensor-to-scalar ratio $r=0.2^{+0.07}_{-0.05}$. This result may be reconciled with the Planck experiment [arXiv:1303.5082] if there is a considerable tilt of r, $\mathcal{T}_r$, with a positive sign, $\mathcal{T}_r=d\ln r/d\ln k\sim 0.16$ corresponding to a blue tilt for the tensor modes of order $n_T\simeq0.12$. Simple slow-roll models fail to provide this tilt. In this note we show that a non-Bunch-Davis initial state for perturbations not only provides a match between large field chaotic models (like $m^2\phi^2$) with the Planck result [arXiv:1306.4914], but also naturally accommodates the running of the tilt needed to explain the BICEP2 result.
\end{document}