BMS 3 fermionic localization

Joan Simón, and Boyang Yu · 2024

Read BMS 3 fermionic localization by Joan Simón, and Boyang Yu (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-05038.txt.

Introduction

Joan Simon and Boyang Yu compute an exact value for a gravity model in three dimensions in a 2024 paper. The model is pure gravity with zero cosmological constant, written as a geometric action. Using a method called fermionic localization, they work out the partition function for a torus shape. The case is one fixed orbit of the BMS symmetry group. The result lets them discuss whether the one loop level of the calculation is already exact.

Catalog record

CREATORJoan Simón, and Boyang Yu
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE67 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-05038

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

BMS 3 fermionic localization Joan Simón, and Boyang Yu Abstract We consider the geometric action formulation for 3d pure gravity with vanishing cosmological constant. We use fermionic localization to compute the exact torus partition function for a constant representative coadjoint orbit of BMS ^ 3 \widehat{\text{BMS}}_{3} . This allows us to discuss its 1-loop exactness. 1 Introduction Much progress has been achieved in the understanding of quantum gravity in low spacetime dimensions due to the development of quantization methods. For example, leveraging the Chern-Simons (CS) formulation of AdS 3 pure gravity enables a rigorous exploration of boundary conditions, gauge fixing, and the presence of non-trivial holonomies in the CS connection Achucarro:1986uwr ; Witten:1988hc . After performing the Hamiltonian reduction, the Chern-Simons action reduces to a Wess-Zumino-Witten (WZW) boundary action Elitzur:1989nr ; Coussaert:1995zp . Simultaneously, Alekseev and Shatashvili applied the coadjoint orbit method to study the Virasoro group Alekseev:1988ce and reproduced the same action by deforming the geometric action associated to the coadjoint orbit with a Hamiltonian Alekseev:1990mp . Since the coadjoint orbits of the Virasoro group form symplectic spaces Kirillov , these can be quantized using geometric quantization Witten:1987ty or phase space path integrals Alekseev:1988vx . This connection led to the coadjoint orbit quantization of AdS 3 Cotler:2018zff . In recent years, such connection was generalized to 3d Minkowski Barnich:2014kra ; Barnich:2015uva ; Oblak:2016eij ; Barnich:2017jgw , dS 3 Cotler:2019nbi , AdS 3 with Compère-Song-Strominger boundary conditions Compere:2013bya ; Detournay:2024gth or Rindler boundary conditions Afshar:2024ess . This quantization technology allows to compute quantum effects in these gravitational theories 1 1 1 Even though the classical action of Chern-Simons action coincides with the three dimensional Einstein gravity action, the quantization of both theories is different Witten:1988hc . For example, metrics should be invertible in gravity, whereas there is no analogous requirement in the Chern-Simons gauge theory. As a result, the Chern-Simons path integral will include field configurations which cannot be interpreted as three-dimensional metrics. However, if one focuses on perturbations around a sensible (invertible) metric, the Chern-Simons theory still provides a valid description of the gravity theory. The latter is the approach followed here. . This includes loop corrections to partition functions Maloney:2007ud ; Giombi:2008vd , correlation functions Bhatta:2016hpz ; Cotler:2018zff ; Merbis:2019wgk or even entanglement entropy measures Ammon:2013hba ; deBoer:2013vca ; Castro:2014tta ; Bagchi:2014iea ; Basu:2015evh . A further development is the 1-loop exactness of the path integrals over these geometric actions under some conditions. This was established for the Schwarzian theory Stanford:2017thb , AdS 3 gravity Cotler:2018zff and the BMS 2 Schwarzian theory Afshar:2021qvi . The purpose of this note is to explore the geometric action formulation for 3d pure Einstein gravity with vanishing cosmological constant to ask whether the one-loop contribution to the torus partition function is exact 2 2 2 Perturbative computations can be found in Barnich:2015mui ; Garbarz:2015lua ; Merbis:2019wgk ; Leston:2023ugd . . To answer this question, we use fermionic localization. Localization is a powerful technique to compute exact quantities in supersymmetric quantum field theories Duistermaat:1982vw ; Atiyah:1984px ; Witten:1988ze , such as partition functions, Wilson loops, and other observables in several dimensions, including applications to gauge theories, string theory, and black hole entropy Pestun:2007rz ; Hama:2011ea ; Closset:2012ru ; Benini:2015eyy ; Iliesiu:2022kny . The key point is to localize the path integral to a finite-dimensional subset of field configurations. This is achieved by deforming the action with a Q-exact term QF , where Q is a supersymmetry generator, that ensures the integral is dominated by the fixed points of QF . This localization method has also been applied to the Schwarzian theory Stanford:2017thb and AdS 3 gravity Cotler:2018zff relying on their phase space being K \ddot{\text{a}} hler. Even though it is not known to us whether such structure exists for the phase space of \widehat{\text{BMS}}_{3} , we will construct such QF -term using field theory techniques, enabling us to perform the path integral exactly and to probe the one-loop exactness of the torus partition function for BMS 3 gravity. This paper is organized as follows. In section 2 , we briefly review the fermionic localization technology used in the main text. In section 3 , we rederive the 1-loop exactness for the Virasoro partition function using the same logic and tools that we apply later in section 4 to compute the torus BMS 3 partition function. In section 5 , we summarize our results. The supersymmetry of the geometric actions used to perform our calculations is discussed in appendix A . Note Added: While finishing our work, reference Cotler:2024cia appeared. As part of the results presented in Cotler:2024cia , it is also claimed, and shown, that the one-loop partition function of 3d pure gravity with a vanishing cosmological constant is 1-loop exact around the Minkowski vacuum. Our results are consistent, but derived using fermionic localization rather than performing the direct path integral by observing the linear functional dependence on the superrotation variable, and applicable to other gravitational saddles, such as conical defects and flat space cosmologies. 2 Fermionic localization The technique we use to compute the torus partition function is the method of fermionic localization. This is briefly reviewed below following Cotler:2018zff . Given a symplectic phase space \mathcal{M} 3 3 3 In this note, \mathcal{M} will correspond to the coadjoint orbit of the asymptotic symmetry G preserving a set of gravitational boundary conditions. In physics terminology, this corresponds to the phase space of physical configurations connected to a given classical saddle by the set of large gauge transformations preserving the asymptotic boundary conditions. , the quantization of the classical theory leads to the partition function \mathcal{Z}=\int_{\mathcal{M}}[dx^{i}]\,\text{Pf}(\omega)\,e^{-S_{\textrm{\tiny E}}}\,, (1) where \text{Pf}(\omega) is the Pfaffian of the symplectic form \omega in \mathcal{M} and S_{\textrm{\tiny E}} is the Euclidean action resulting from the Wick rotation t\to-iy of the Alekseev-Shatashvili (AS) type action 4 4 4 In this note, the geometric action ( 2 ) will correspond to the 3d pure gravity bulk action together with a boundary term to have a well defined variational principle. Alekseev:1990mp S_{\textrm{\tiny geometric}}=\int_{\gamma}\left(a+H_{v}\right)\,dt\,, (2) with \omega=\delta a and H v the Noether charge associated with the global symmetry generated by the flow of v, i.e. i_{v}\omega=dH_{v} . Explicitly, S_{\textrm{\tiny E}}=\int dy\left(i\frac{\partial x^{i}}{\partial y}a_{i}+H\right). (3) Writing the Pfaffian term as an integral over the Grassmann-odd ghost fields \psi^{i} , the resulting path integral becomes \mathcal{Z}=\int[dx^{i}][d\psi^{i}]\,e^{-S_{\textrm{\tiny E}}^{\prime}},\qquad\text{with}\quad S_{\textrm{\tiny E}}^{\prime}\equiv S_{\textrm{\tiny E}}+S_{\omega}=S_{\textrm{\tiny E}}-\frac{1}{2}\int dy\,\omega_{ij}\psi^{i}\psi^{j}. (4) The total action S_{\textrm{\tiny E}}^{\prime} is invariant under the symmetry generated by the Grassmann-odd supercharge Q whose actions on the dynamical fields is Qx^{i}=\psi^{i},\quad Q\psi^{i}=V^{i}\equiv-\omega^{ij}\frac{\delta S_{\textrm{\tiny E}}}{\delta x^{i}}\,. (5) According to the Duistermaat-Heckman theorem Duistermaat:1982vw , the integral of a function which is both Q-exact and Q-closed vanishes. As a consequence, deforming S_{\textrm{\tiny E}}^{\prime}\to S_{\textrm{\tiny E}}^{\prime}+sQF in ( 4 ), with QF satisfying Q^{2}F=0\qquad\text{and}\qquad(QF)_{\textrm{\tiny bosonic}}\geq 0, (6) does not modify the path integral, i.e. \mathcal{Z}=\mathcal{Z}[s]:=\int[dx^{i}][d\psi^{i}]\,e^{-(S^{\prime}_{\textrm{\tiny E}}+sQF)} (7) Since the s\to\infty limit localizes the path integral to the localization manifold \mathcal{M}_{\textrm{\tiny loc}}=\{x_{c}^{i}\,\,|\,\,(QF)_{\textrm{\tiny bosonic}}[x_{c}]=0,\psi^{i}=0\}\,, (8) and all higher loop contributions are suppressed compared to the one-loop term, one reaches the conclusion Banerjee:2009af ; Dabholkar:2010uh \mathcal{Z}=\lim_{s\to\infty}\mathcal{Z}[s]_{\textrm{\tiny 1-loop}}=\int_{\mathcal{M}_{\textrm{\tiny loc}}}[dx_{c}]\,e^{-S_{\textrm{\tiny E}}[x_{c}]}\frac{1}{\text{SDet}^{\prime}(QF)_{x_{c}}} (9) where SDet is the superdeterminant given by the ratio of the bosonic and fermionic determinants at 1-loop and the prime means the

Read the complete text