Associators for AdS string amplitude building blocks

Konstantin Baune · 2025

Read Associators for AdS string amplitude building blocks by Konstantin Baune (2025) 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-2505-23385.txt.

Introduction

Konstantin Baune shows that the building blocks of string amplitudes in curved anti-de Sitter space grow from a single seed object, in a 2025 paper. For open strings the generator is the Drinfeld associator; for closed strings, the Deligne associator. The result lifts known flat-space recursions into the curved setting. It also proves the blocks expand at low energy with multiple zeta values as coefficients, the closed-string ones being the single-valued kind. The paper supplies all-order relations among the integral forms of the blocks.

Catalog record

CREATORKonstantin Baune
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE67 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-23385

Browse the full library · About the Tower of Babel

Full text

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

Associators for AdS string amplitude building blocks Konstantin Baune Abstract We show that building blocks for open- and closed-string amplitudes on AdS are generated by the Drinfeld and Deligne associator, respectively. Our formalism lifts the known associator recursions for flat-space string amplitudes to the AdS picture. This delivers another proof that the AdS building blocks admit low-energy expansions with (single-valued) multiple zeta values as coefficients and provides all-order relations for the integral expressions. 1 Introduction Scattering amplitudes in quantum field theory and string theory are the simplest observables, whose analytic and algebraic properties need to be understood. When considering possible four-point amplitudes with infinite towers of mediated particles, Veneziano wrote down the Veneziano amplitude as the Euler beta function [ 1 ] \beta(s,t)=\int_{0}^{1}dx\,x^{s-1}(1-x)^{t-1}=\frac{\Gamma(s)\Gamma(t)}{\Gamma(s+t)}, (1.1) which led to the birth of string theory and describes the four-point open-string tree-level amplitude. Later, also the four-point expression for the closed string was found, known as the Virasoro–Shapiro amplitude , which can be written as the complex beta function [ 2 , 3 ] \beta_{\mathbb{C}}(s,t)=-\frac{1}{2\pi i}\int_{\mathbb{C}}dz\,d\overline{z}\,|z|^{2(s-1)}|1-z|^{2(t-1)}=\frac{\Gamma(s)\Gamma(t)\Gamma(1-s-t)}{\Gamma(s+t)\Gamma(1-s)\Gamma(1-t)}. (1.2) Comparing the two integral expression for \beta and \beta_{\mathbb{C}} , one might suspect a relation: The most well-known identity between them is the Kawai–Lewellen–Tye (KLT) relation [ 4 ] , which allows to express n -point closed-string tree-level amplitudes as “squares” of n -point open-string amplitudes. The four-point KLT relation reads [ 4 , 5 ] \beta_{\mathbb{C}}(s,t)=-\frac{1}{2\pi i}\left(\frac{2\sin(\pi s)\sin(\pi t)}{i\,\sin(\pi(s+t))}\right)\beta(s,t)^{2}. (1.3) This identity was later related to intersection theory [ 6 , 7 ] and also generalized for one-loop string amplitudes [ 8 , 9 , 10 , 11 ] . Another way to formulate the relation between open- and closed-string amplitudes is through the single-valued map . The simplest installation becomes visible when considering the low-energy expansions (i.e. small Mandelstam variables s,\,t ) of the above amplitudes (see e.g. refs. [ 12 , 5 ] ) \displaystyle\beta(s,t) \displaystyle=\left(\frac{1}{s}+\frac{1}{t}\right)\exp\Bigg[-\sum_{n\geq 2}\frac{\zeta(n)}{n}\left(s^{n}+t^{n}-(s+t)^{n}\right)\Bigg], (1.4a) \displaystyle\beta_{\mathbb{C}}(s,t) \displaystyle=\left(\frac{1}{s}+\frac{1}{t}\right)\exp\Bigg[-\sum_{\begin{subarray}{c}n\geq 2\\ n\text{ odd}\end{subarray}}\frac{2\zeta(n)}{n}\left(s^{n}+t^{n}-(s+t)^{n}\right)\Bigg], (1.4b) where one notices that \zeta(2n)\mapsto 0,\quad\zeta(2n{-}1)\mapsto 2\zeta(2n{-}1),\quad n\in\mathbb{N}, (1.5) maps \beta(s,t) to \beta_{\mathbb{C}}(s,t) . These are simple instances of the single-valued map for multiple zeta values, established in refs. [ 13 , 14 ] and connected to tree-level string amplitudes of all multiplicities in refs. [ 12 , 15 ] . On the level of the full integrals, this relation was proven by the means of single-valued integration [ 16 , 5 ] , so that the integral \beta_{\mathbb{C}}(s,t) is the single-valued image of \beta(s,t) through a mapping of twisted (co)homologies. In recent years, the connection of string amplitudes to polylogarithms and the Knizhnik–Zamolodchikov equation was studied thoroughly. This included finding that the Drinfeld associator, which can be represented as the generating series of multiple zeta values, recursively generates all tree-level open-string amplitudes [ 17 , 18 , 19 ] . Correspondingly, the Deligne associator, as the generating series of all single-valued multiple zeta values, recursively generates the closed-string tree-level amplitudes [ 12 , 15 , 5 , 20 ] . While the structure of tree-level amplitudes in flat-space string theory is by now well understood, the interest in recent years was extended to considering string-amplitudes on AdS backgrounds. This was first established through a bootstrap program to find the large AdS radius corrections to the Virasoro–Shapiro amplitude on AdS_{5}\times S^{5} [ 21 , 22 , 23 , 24 , 25 ] . It was continued for other theories [ 26 , 27 , 28 , 29 ] as well as for the AdS corrections to the Veneziano amplitude [ 30 , 31 ] . One might wonder, how the KLT relations and the Drinfeld and Deligne recursions found in flat space generalize when considering the corresponding amplitudes on AdS. For this, one has to consider the types of integrals appearing for the curvature corrections, which are (complex) beta function integrals with insertions of (single-valued) polylogarithms in the integrand. In ref. [ 31 ] it was already shown, that in the open-string case, the low-energy expansion coefficients are zeta values. Correspondingly in ref. [ 23 ] , it was proven that the coefficients are single-valued zeta values in the closed-string case. Very recently, ref. [ 32 ] found that these integrals also obey double-copy relations, generalizing the KLT relation to the AdS integrals. In ref. [ 32 ] the question was raised, which types of associators correspond to AdS string amplitudes and their building blocks in the open- and closed-string cases. The current article aims to answer this question. Summary. We will show how the AdS amplitude building blocks for open and closed strings can be obtained through recursions using the Drinfeld associator and the Deligne associator, respectively, in suitable matrix representations. This yields all order expansions of the building blocks in the low-energy limit of small Mandelstam variables. Additionally, it provides another proof that the coefficients of these expansions are multiple zeta values for open strings and single-valued multiple zeta values for closed strings. Furthermore, our construction indicates that similar to flat space, Drinfeld and Deligne associators generate four-point open- and closed-string amplitudes on AdS. Finally, we will present how these structures generalize for iterated integrals with polylogarithm insertions, reminiscent of higher-point amplitudes. Outline. The article is structured as follows. In Section 2 we review string integral recursions using the Drinfeld and Deligne associators, respectively. An introduction to four-point AdS amplitude building blocks for open and closed strings is provided in Section 3 . In Section 4 we are going to present and study the new associator relations for the AdS building blocks of open and closed strings. Finally, we will discuss further research directions in Section 5 . Three appendices supplement this article: In Appendix A we summarize the basics and fix our conventions for multiple polylogarithms, multiple zeta values and their single-valued analogs, as well as the Drinfeld and Deligne associator. Appendix B collects results for the open-string Drinfeld associator relations of weight three AdS amplitude building blocks, while Appendix C contains details for a higher-point example. 2 Drinfeld and Deligne recursions The Drinfeld associator \Phi and the Deligne associator \Phi^{\operatorname{sv}} —as reviewed in Appendix A —are the generating series of multiple zeta values (MZVs) and single-valued multiple zeta values (svMZVs), respectively, \displaystyle\Phi(e_{0},e_{1}) \displaystyle=\sum_{w\in\{e_{0},e_{1}\}^{\times}}(-1)^{\#w}w\,\zeta_{w}, (2.1a) \displaystyle\Phi^{\operatorname{sv}}(e_{0},e_{1}) \displaystyle=\sum_{w\in\{e_{0},e_{1}\}^{\times}}(-1)^{\#w}w\,\zeta^{\operatorname{sv}}_{w}=\operatorname{sv}(\Phi). (2.1b) They are valued in the free algebra generated by the letters e_{0},\,e_{1} , where \{e_{0},e_{1}\}^{\times} refers to the set of all words in e_{0} and e_{1} including the empty word e and we have the natural mapping \zeta_{w}=\zeta_{\widetilde{w}} with w\,{=}\,\prod_{j=1}^{k}e_{0}^{a_{j}}e_{1}^{b_{j}}\mapsto\widetilde{w}\,{=}\,\prod_{j=1}^{k}0^{a_{j}}1^{b_{j}} for (sv)MZVs. The expression \#w in eq. ( 2.1 ) refers to the number of e_{1} ’s appearing in a word w . Our conventions for polylogarithms and multiple zeta values agree with ref. [ 32 ] and are collected in Appendix A . Let us furthermore define \mathcal{Z}=\mathbb{Q}[(\zeta_{w})_{w\in\{0,1\}^{\times}}] and \mathcal{Z}^{\operatorname{sv}}=\mathbb{Q}[(\zeta_{w}^{\operatorname{sv}})_{w\in\{0,1\}^{\times}}]\subset\mathcal{Z} as the rings over \mathbb{Q} generated by MZVs and svMZVs, respectively. The above associators can be used to recursively generate and calculate integrals of Selberg type, such as string amplitudes, as follows. Let F(x) be a meromorphic vector of iterated integrals, solving the Knizhnik–Zamolodchikov (KZ) equation \frac{d}{dx}F(x)=\left(\frac{e_{0}}{x}+\frac{e_{1}}{x-1}\right)F(x), (2.2) for square matrices e_{0},\,e_{1} . Then the regularized boundary values C_{0}=\lim_{x\to 0}x^{-e_{0}}F(x),\qquad C_{1}=\lim_{x\to 1}(1-x)^{-e_{1}}F(x) (2.3) are related by the Drinfeld associator via (see e.g. [ 33 ] ) C_{1}=\Phi(e_{0},e_{1})\,C_{0}, (2.4)

Read the complete text