Violation of Energy Bounds in Designer Gravity
Thomas Hertog
TL;DR
The paper addresses the stability of designer gravity theories—AdS gravity with tachyonic scalars under boundary conditions defined by a function $W$—and questions a proposed energy bound tied to the minimum of $W$ when the scalar potential derives from a superpotential $P$. By analyzing asymptotics, spinor charges, and taking conformal rescalings of AdS-invariant solitons, the authors construct explicit counterexamples where negative-energy initial data exist even for $V$ arising from $P$ and $eta(\alpha)$ determined by $W$. The results show that the positive-energy guarantees depend critically on both $P$ and the boundary data; only certain combinations yield a stable ground state, while others admit arbitrarily negative mass solitons. In the AdS/CFT context, these findings imply that while many boundary deformations preserve positivity, a universal PET does not hold for all designer-gravity boundary conditions, and criteria linking $P$ and $W$ must be identified for spinor-based positivity to apply.
Abstract
We continue our study of the stability of designer gravity theories, where one considers anti-de Sitter gravity coupled to certain tachyonic scalars with boundary conditions defined by a smooth function W. It has recently been argued there is a lower bound on the conserved energy in terms of the global minimum of W, if the scalar potential arises from a superpotential P and the scalar reaches an extremum of P at infinity. We show, however, there are superpotentials for which these bounds do not hold.
