Laurent type expansion of multiple polylogarithms at integer points

Pawan Singh Mehta and Biswajyoti Saha · 2026

Read Laurent type expansion of multiple polylogarithms at integer points by Pawan Singh Mehta and Biswajyoti Saha (2026) 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-2601-17539.txt.

Introduction

Pawan Singh Mehta and Biswajyoti Saha study multiple polylogarithms near integer points. Polylogarithms are special functions with deep ties to number theory. Their local behavior in the s-aspect is the focus. The authors write a Laurent type expansion at integer points. The expansion uses power series and rational functions.

Local expansions let mathematicians compute values near hard points. They also expose poles and singular behavior in a clear form. Multiple polylogarithms appear in zeta values and particle physics. A firm expansion serves both fields. The work adds precision tools for a classic function family. It can speed proofs that need values at integers.

Catalog record

CREATORPawan Singh Mehta and Biswajyoti Saha
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE98 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-17539

Browse the full library · About the Tower of Babel

Full text

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

Laurent type expansion of multiple polylogarithms at integer points Pawan Singh Mehta and Biswajyoti Saha Abstract . In this article, we study the local behaviour of the multiple polylogarithm functions at integer points, in the s s -aspect. This is done by writing a Laurent type expansion at integer points, involving certain power series and rational functions. The coefficients of these power series are the regularised values of the multiple polylogarithm functions at certain related integer points. 2020 Mathematics Subject Classification 11M32, 32Dxx 1. Introduction Throughout this article, we denote the set of all non-negative integers by {\mathbb{N}} . For positive integers r , a_{1},\ldots,a_{r} and complex numbers z_{1},\ldots,z_{r} with |z_{i}|<1 for all 1\leq i\leq r , the multiple polylogarithm (of depth r ) is defined by the convergent power series \mathrm{Li}(a_{1},\ldots,a_{r};z_{1},\ldots,z_{r}):=\sum_{n_{1}>\cdots>n_{r}>0}\frac{z_{1}^{n_{1}}\cdots z_{r}^{n_{r}}}{n_{1}^{a_{1}}\cdots n_{r}^{a_{r}}}. This definition can be extended for |z_{i}|\leq 1 for all 1\leq i\leq r , if a_{1}\geq 2 . Further, this series can be considered as a multiple Dirichlet series by fixing complex numbers z_{1},\ldots,z_{r} and replacing a_{1},\ldots,a_{r} by complex variables s_{1},\ldots,s_{r} such that (s_{1},\ldots,s_{r})\in U_{r} , i.e., (1) \mathrm{Li}_{\bf z}(s_{1},\ldots,s_{r})=\mathrm{Li}_{(z_{1},\ldots,z_{r})}(s_{1},\ldots,s_{r}):=\sum_{n_{1}>\cdots>n_{r}>0}\frac{z_{1}^{n_{1}}\cdots z_{r}^{n_{r}}}{n_{1}^{s_{1}}\cdots n_{r}^{s_{r}}}, where U_{r}:=\{(s_{1},\ldots,s_{r})\in{\mathbb{C}}^{r}~:~\Re(s_{1}+\cdots+s_{i})>i~\text{ for all }~1\leq i\leq r\}. The above series converges normally on compact subsets of U_{r} and therefore defines a holomorphic function there (see [ 5 , Proposition 2] ). The depth 0 multiple polylogarithm function is taken to be the constant function 1 . In our recent work [ 4 ] , we have shown that the series in ( 1 ) converges (conditionally) on a larger open domain, depending on {\bf z}=(z_{1},\ldots,z_{r}) , by treating it as (2) \lim_{N\to\infty}\sum_{N>n_{1}>\cdots>n_{r}>0}\frac{z_{1}^{n_{1}}\cdots z_{r}^{n_{r}}}{n_{1}^{s_{1}}\cdots n_{r}^{s_{r}}}. For an integer r\geq 1 and {\bf z}=(z_{1},\ldots,z_{r})\in{\mathbb{C}}^{r} , where |z_{i}|\leq 1 for all 1\leq i\leq r , let q({\bf z}) be the smallest positive integer such that 1\leq q({\bf z})\leq r and z_{q({\bf z})}\neq 1 . If no such q({\bf z}) exists, then we set q({\bf z}):=r+1 . Consider the open set U_{r}(\mathbf{z}) of \mathbb{C}^{r} defined as follows: (3) \begin{split}U_{r}(\mathbf{z}):=\{(s_{1},\ldots,s_{r})\in\mathbb{C}^{r}:&\ \Re(s_{1}+\cdots+s_{i})>i\text{ if }1\leq i<q({\bf z})\\ &\text{ and }\Re(s_{1}+\cdots+s_{i})>i-1\text{ if }q({\bf z})\leq i\leq r\}.\end{split} We proved the following in [ 4 , Theorem 2] . Theorem 1 . The multiple Dirichlet series in ( 1 ) converges on U_{r}(\mathbf{z}) . Moreover, the series in ( 1 ) defines a holomorphic function on U_{r}(\mathbf{z}) . This theorem therefore extends the classical result for the depth 1 Dirichlet series \mathrm{Li}_{(z)}(s) . However, this open domain of convergence is not optimal in general. For example, as N\to\infty , one has (4) \sum_{N>n_{1}>n_{2}>0}\frac{(-1)^{n_{2}}}{n_{1}^{2}n_{2}^{-1}}=\frac{\log 2}{2}-\frac{\pi^{2}}{16}+o(1). Clearly, (2,-1)\notin U_{2}(1,-1) . In fact, one can see that for s\in{\mathbb{C}} with \Re(s)>1 , the limit \lim_{N\to\infty}\sum_{N>n_{1}>n_{2}>0}\frac{(-1)^{n_{2}}}{n_{1}^{s}n_{2}^{-1}} exists. In [ 4 , Theorem 3] , we established a more accurate and larger open domain where the series in ( 1 ) (viewed as the limit in ( 2 )) converges at integer points. To recall the statement, we need a few more notations. Let r\geq 1 be an integer and \mathbf{z}=(z_{1},\ldots,z_{r})\in{\mathbb{C}}^{r} be such that z_{1},\ldots,z_{r} are roots of unity. For each 1\leq i\leq r , define (5) J_{i}(\mathbf{z}):=\{j:1\leq j\leq i\text{ and }z_{[j,i]}=1\},\ \text{ and }\ Q_{i}(\mathbf{z}):=|J_{i}(\mathbf{z})|, the number of elements in J_{i}(\mathbf{z}) , where z_{[j,i]}:=\prod_{k=j}^{i}z_{k} for 1\leq j\leq i\leq r . Consider the open set (6) V_{r}(\mathbf{z}):=\{(s_{1},\ldots,s_{r})\in{\mathbb{C}}^{r}:\Re(s_{1}+\cdots+s_{i})>Q_{i}(\mathbf{z})\text{ for all }1\leq i\leq r\}. It is easy to see that V_{r}(\mathbf{z})=U_{r} if and only if \mathbf{z}=(1,\ldots,1) . Moreover, U_{r}(\mathbf{z})\subseteq V_{r}(\mathbf{z}) . We proved the following in [ 4 , Theorem 3] . Theorem 2 . Let r\geq 1 be an integer and z_{1},\ldots,z_{r}\in{\mathbb{C}} be roots of unity. The multiple Dirichlet series in ( 1 ) converges at integer points of V_{r}(\mathbf{z}) . Moreover, for an integer point (s_{1},\ldots,s_{r})\in V_{r}(\mathbf{z}) and k_{1},\ldots,k_{r}\in\mathbb{N} , the series (7) \sum_{n_{1}>\cdots>n_{r}>0}\frac{z_{1}^{n_{1}}(\log n_{1})^{k_{1}}\cdots z_{r}^{n_{r}}(\log n_{r})^{k_{r}}}{n_{1}^{s_{1}}\cdots n_{r}^{s_{r}}} converges when we consider it as the limit (8) \lim_{N\to\infty}\sum_{N>n_{1}>\cdots>n_{r}>0}\frac{z_{1}^{n_{1}}(\log n_{1})^{k_{1}}\cdots z_{r}^{n_{r}}(\log n_{r})^{k_{r}}}{n_{1}^{s_{1}}\cdots n_{r}^{s_{r}}}. This theorem was proved by devising a regularisation process for the series defining the multiple polylogarithm functions (see [ 4 , §2] ). However, the question of holomorphicity of the multiple polylogarithm functions at integer points of V_{r}(\mathbf{z}) is a delicate one and was not addressed in [ 4 ] . In this article, which is a follow-up of [ 4 ] , we use the regularisation process devised in our article [ 4 ] to address this problem by writing explicit Laurent type expansion for the multiple polylogarithm functions around all the integer points in {\mathbb{C}}^{r} , for any positive integer r . A prototype of the desired Laurent type expansion is \zeta(s)=\frac{1}{s-1}+\sum_{k\geq 0}\frac{(-1)^{k}\gamma_{k}}{k!}(s-1)^{k}, where for k\geq 0 , \gamma_{k} is the Stieltjes constant defined by \gamma_{k}:=\lim_{N\to\infty}\left(\sum_{N>n>0}\frac{\log^{k}n}{n}-\frac{\log^{k+1}N}{k+1}\right). In [ 6 ] , the second author provided explicit Laurent type expansion for the multiple zeta functions around integer points. Our work here therefore extends the work of [ 6 ] to the case of the multiple polylogarithm functions. This requires us to study the regularised multiple polylogarithm functions, which we define as follows: For a positive integer r , consider {\bf a}=(a_{1},\ldots,a_{r})\in{\mathbb{Z}}^{r} and {\bf z}=(z_{1},\ldots,z_{r})\in{\mathbb{C}}^{r} , such that z_{1},\ldots,z_{r} are roots of unity. Then we consider the formal power series (9) \displaystyle\sum_{k_{1},\ldots,k_{r}\geq 0}\frac{(-1)^{k_{1}+\cdots+k_{r}}}{k_{1}!\cdots k_{r}!}\ell_{[k_{1},\ldots,k_{r}]}^{(a_{1},\ldots,a_{r})}{(\mathbf{z})}(s_{1}-a_{1})^{k_{1}}\cdots(s_{r}-a_{r})^{k_{r}}, where the constant \ell_{[k_{1},\ldots,k_{r}]}^{(a_{1},\ldots,a_{r})}(\mathbf{z}) is the regularised value of the sequence of partial sums (10) \displaystyle\sum_{N>n_{1}>\cdots>n_{r}>0}\frac{z_{1}^{n_{1}}(\log n_{1})^{k_{1}}\cdots z_{r}^{n_{r}}(\log n_{r})^{k_{r}}}{n_{1}^{a_{1}}\cdots n_{r}^{a_{r}}}, as defined in [ 4 , Definitions 2, 3] . For the sake of completeness, we recall from [ 4 ] that as N\to\infty , the series ( 10 ) can be expressed as a sum of four different types of terms, namely, one which goes to \infty in modulus, one which is bounded but oscillates, one which remains constant and one which goes to 0 . The constant term is then defined to be the regularised value \ell_{[k_{1},\ldots,k_{r}]}^{(a_{1},\ldots,a_{r})}(\mathbf{z}) of the series ( 10 ). For example, \sum_{N>n>0}(-1)^{n}=-\frac{(-1)^{N}}{2}-\frac{1}{2},\ \sum_{N>n>0}(-1)^{n}n=-\frac{(-1)^{N}N}{2}+\frac{(-1)^{N}}{4}-\frac{1}{4}, and further, as N\to\infty , we also have \sum_{N>n_{1}>n_{2}>0}\frac{(-1)^{n_{2}}n_{2}}{n_{1}}=-\frac{\log N}{4}+\frac{(-1)^{N}}{4}+\frac{1-\log 2-\gamma}{4}+o(1), where \gamma is the Euler’s constant. Thus, \ell_{[0]}^{(0)}(-1)=-1/2 , \ell_{[0]}^{(-1)}(-1)=-1/4 and \ell_{[0,0]}^{(1,-1)}(1,-1)=(1-\log 2-\gamma)/4 . For more details on the above regularisation process we suggest the reader to see [ 4 , §2] . Now for the formal power series ( 9 ), we prove that it converges in a neighbourhood of {\bf a} and it can be extended to a meromorphic function on {\mathbb{C}}^{r} . We denote this meromorphic function by \mathrm{Li}^{\text{Reg}}_{(\mathbf{z};\mathbf{a})}(s_{1},\ldots,s_{r}) and call it the (depth r ) regularised multiple polylogarithm function for \mathbf{z} and \mathbf{a} . The depth 0 regularised multiple polylogarithm function is taken to be the constant function 1 . If {\bf a}\in{\mathbb{Z}}^{r}\cap U_{r}(\mathbf{z}) , then as meromorphic functions on {\mathbb{C}}^{r} we have \mathrm{Li}_{\mathbf{z}}(s_{1},\ldots,s_{r})=\mathrm{Li}^{\text{Reg}}_{(\mathbf{z};\mathbf{a})}(s_{1},\ldots,s_{r}). But for other integer points of {\mathbb{C}}^{r} , this is not necessarily true. For example, if r=2 and \mathbf{z}=(1,-1) , then around the point \mathbf{a}=(1,1) , we have the following equality between meromorphic functions: (11) \displaystyle\mathrm{Li}_{\mathbf{z}}(s_{1},s_{2})=\frac{\mathrm{Li}^{\text{Reg}}_{(-1;1)}{(s_{2})}}{s_{1}-1}+\mathrm{Li}^{\text{Reg}}_{(\mathbf{z};\mathbf{a})}(s_{1},s_{2}). Hence s_{1}=1 is the only polar hyperplane of \mathrm{Li}_{\mathbf{z}}(s_{1},s_{2}) which passes through the point (1,1) . The formulation of our theorem is more delicate for general integer points, particularly for those lying outside \overline{U_{r}(\mathbf{z})} . Hence we first consider the case of integer points in \overline{U_{r}(\mathbf{z})} . This has been stated in Theorem 4 in § 2 and the proof is given in § 4 . The intermediate steps involved in the proof of Theorem 4 are discussed in

Read the complete text