-
Learning the symmetric group: large from small
Authors:
Max Petschack,
Alexandr Garbali,
Jan de Gier
Abstract:
Machine learning explorations can make significant inroads into solving difficult problems in pure mathematics. One advantage of this approach is that mathematical datasets do not suffer from noise, but a challenge is the amount of data required to train these models and that this data can be computationally expensive to generate. Key challenges further comprise difficulty in a posteriori interpre…
▽ More
Machine learning explorations can make significant inroads into solving difficult problems in pure mathematics. One advantage of this approach is that mathematical datasets do not suffer from noise, but a challenge is the amount of data required to train these models and that this data can be computationally expensive to generate. Key challenges further comprise difficulty in a posteriori interpretation of statistical models and the implementation of deep and abstract mathematical problems.
We propose a method for scalable tasks, by which models trained on simpler versions of a task can then generalize to the full task. Specifically, we demonstrate that a transformer neural-network trained on predicting permutations from words formed by general transpositions in the symmetric group $S_{10}$ can generalize to the symmetric group $S_{25}$ with near 100\% accuracy. We also show that $S_{10}$ generalizes to $S_{16}$ with similar performance if we only use adjacent transpositions. We employ identity augmentation as a key tool to manage variable word lengths, and partitioned windows for training on adjacent transpositions. Finally we compare variations of the method used and discuss potential challenges with extending the method to other tasks.
△ Less
Submitted 18 February, 2025;
originally announced February 2025.
-
Rational symmetric functions from the Izergin-Korepin 19-vertex model
Authors:
Alexandr Garbali,
Weiying Guo,
Michael Wheeler
Abstract:
Starting from the Izergin-Korepin 19-vertex model in the quadrant, we introduce two families of rational multivariate functions $F_S$ and $G_S$; these are in direct analogy with functions introduced by Borodin in the context of the higher-spin 6-vertex model in the quadrant.
We prove that $F_S(x_1,\dots,x_N;z)$ and $G_S(y_1,\dots,y_M;z)$ are symmetric functions in their alphabets…
▽ More
Starting from the Izergin-Korepin 19-vertex model in the quadrant, we introduce two families of rational multivariate functions $F_S$ and $G_S$; these are in direct analogy with functions introduced by Borodin in the context of the higher-spin 6-vertex model in the quadrant.
We prove that $F_S(x_1,\dots,x_N;z)$ and $G_S(y_1,\dots,y_M;z)$ are symmetric functions in their alphabets $(x_1,\dots,x_N)$ and $(y_1,\dots,y_M)$, and pair together to yield a Cauchy identity. Both properties are consequences of the Yang-Baxter equation of the model.
We show that, in an appropriate limit of the spectral parameters $z$, $F_S$ tends to a stable symmetric function denoted $H_S$. This leads to a simplified version of the Cauchy identity with a fully factorized kernel, and suggests self-duality of the functions $H_S$.
We obtain a symmetrization formula for the function $F_S(x_1,\dots,x_N;z)$, which exhibits its symmetry in $(x_1,\dots,x_N)$. In contrast to the 6-vertex model, where $F^{6{\rm V}}_S(x_1,\dots,x_N;z)$ is cast as a sum over the symmetric group $\mathfrak{S}_N$, the symmetrization formula in the 19-vertex model is over a larger set of objects that we define; we call these objects 2-permutations.
As a byproduct of the proof of our symmetrization formula, we obtain explicit formulas for the monodromy matrix elements of the 19-vertex model in a basis that renders them totally spatially symmetric.
△ Less
Submitted 23 December, 2024;
originally announced December 2024.
-
Symmetric functions from the six-vertex model in half-space
Authors:
Alexandr Garbali,
Jan de Gier,
William Mead,
Michael Wheeler
Abstract:
We study the stochastic six-vertex model in half-space with generic integrable boundary weights, and define two families of multivariate rational symmetric functions. Using commutation relations between double-row operators, we prove a skew Cauchy identity of these functions. In a certain degeneration of the right-hand side of the Cauchy identity we obtain the partition function of the six-vertex…
▽ More
We study the stochastic six-vertex model in half-space with generic integrable boundary weights, and define two families of multivariate rational symmetric functions. Using commutation relations between double-row operators, we prove a skew Cauchy identity of these functions. In a certain degeneration of the right-hand side of the Cauchy identity we obtain the partition function of the six-vertex model in a half-quadrant, and give a Pfaffian formula for this quantity. The Pfaffian is a direct generalization of a formula obtained by Kuperberg in his work on symmetry classes of alternating-sign matrices. One of our families of symmetric functions admits an integral (sum over residues) formula, and we use this to conjecture an orthogonality property of the dual family. We conclude by studying the reduction of our integral formula to transition probabilities of the (initially empty) asymmetric simple exclusion process on the half-line.
△ Less
Submitted 5 October, 2024; v1 submitted 21 December, 2023;
originally announced December 2023.
-
Shuffle algebras, lattice paths and Macdonald functions
Authors:
Alexandr Garbali,
Ajeeth Gunna
Abstract:
We consider partition functions on the $N\times N$ square lattice with the local Boltzmann weights given by the $R$-matrix of the $U_{t}(\widehat{sl}(n+1|m))$ quantum algebra. We identify boundary states such that the square lattice can be viewed on a conic surface. The partition function $Z_N$ on this lattice computes the weighted sum over all possible closed coloured lattice paths with $n+m$ dif…
▽ More
We consider partition functions on the $N\times N$ square lattice with the local Boltzmann weights given by the $R$-matrix of the $U_{t}(\widehat{sl}(n+1|m))$ quantum algebra. We identify boundary states such that the square lattice can be viewed on a conic surface. The partition function $Z_N$ on this lattice computes the weighted sum over all possible closed coloured lattice paths with $n+m$ different colours: $n$ ``bosonic'' colours and $m$ ``fermionic'' colours. Each bosonic (fermionic) path of colour $i$ contributes a factor of $z_i$ ($w_i$) to the weight of the configuration. We show the following. (i) $Z_N$ is a symmetric function in the spectral parameters $x_1\dots x_N$ and generates basis elements of the commutative trigonometric Feigin--Odesskii shuffle algebra. The generating function of $Z_N$ admits a shuffle-exponential formula analogous to the Macdonald Cauchy kernel. (ii) $Z_N$ is a symmetric function in two alphabets $(z_1\dots z_n)$ and $(w_1\dots w_m)$. When $x_1\dots x_N$ are set to be equal to the box content of a skew Young diagram $μ/ν$ with $N$ boxes the partition function $Z_N$ reproduces the skew Macdonald function $P_{μ/ν}\left[w-z\right]$.
△ Less
Submitted 15 October, 2024; v1 submitted 11 December, 2023;
originally announced December 2023.
-
A $(q,t)$-deformation of the 2d Toda integrable hierarchy
Authors:
Jean-Emile Bourgine,
Alexandr Garbali
Abstract:
A $(q,t)$-deformation of the 2d Toda integrable hierarchy is introduced by enhancing the underlying symmetry algebra $\mathfrak{gl}(\infty)\simeq \text{q-W}_{1+\infty}$ to the quantum toroidal $\mathfrak{gl}(1)$ algebra. The difference-differential equations of the hierarchy are obtained from the expansion of $(q,t)$-bilinear identities, and two equations refining the 2d Toda equation are found in…
▽ More
A $(q,t)$-deformation of the 2d Toda integrable hierarchy is introduced by enhancing the underlying symmetry algebra $\mathfrak{gl}(\infty)\simeq \text{q-W}_{1+\infty}$ to the quantum toroidal $\mathfrak{gl}(1)$ algebra. The difference-differential equations of the hierarchy are obtained from the expansion of $(q,t)$-bilinear identities, and two equations refining the 2d Toda equation are found in this way. The derivation of the bilinear identities follows from the isomorphism between the Fock representation of level $(2,0)$ of the quantum toroidal $\mathfrak{gl}(1)$ algebra and the tensor product of the q-deformed Virasoro algebra with a $u(1)$ Heisenberg algebra. It leads to identify the $(q,t)$-deformed Casimir with the screening charges of the deformed Virasoro algebra. Due to the non-trivial coproduct, equations of the hierarchy no longer involve a single tau-function, but instead relate a set of different tau functions. We then define the universal refined tau function using the $L$-matrix of the quantum toroidal $\mathfrak{gl}(1)$ algebra and interpret it as the generating function of the deformed tau functions. The equations of the hierarchy, written in terms of the universal refined tau function, combine into two-term quadratic equations similar to the $RLL$ equations.
△ Less
Submitted 24 June, 2024; v1 submitted 31 August, 2023;
originally announced August 2023.
-
Computing the R-matrix of the quantum toroidal algebra
Authors:
Alexandr Garbali,
Andrei Neguţ
Abstract:
We consider the problem of the R-matrix of the quantum toroidal algebra $U_{q,t}(\ddot{\mathfrak{gl}}_1)$ in the Fock representation. Using the connection between the R-matrix $R(u)$ ($u$ being the spectral parameter) and the theory of Macdonald operators we obtain explicit formulas for $R(u)$ in the operator and matrix forms. These formulas are expressed in terms of the eigenvalues of a certain M…
▽ More
We consider the problem of the R-matrix of the quantum toroidal algebra $U_{q,t}(\ddot{\mathfrak{gl}}_1)$ in the Fock representation. Using the connection between the R-matrix $R(u)$ ($u$ being the spectral parameter) and the theory of Macdonald operators we obtain explicit formulas for $R(u)$ in the operator and matrix forms. These formulas are expressed in terms of the eigenvalues of a certain Macdonald operator which completely describe the functional dependence of $R(u)$ on the spectral parameter $u$. We then consider the geometric R-matrix (obtained from the theory of K-theoretic stable bases on moduli spaces of framed sheaves), which is expected to coincide with $R(u)$ and thus gives another approach to the study of the poles of the R-matrix as a function of u.
△ Less
Submitted 16 December, 2021;
originally announced December 2021.
-
Shuffle algebras, lattice paths and the commuting scheme
Authors:
Alexandr Garbali,
Paul Zinn-Justin
Abstract:
The commutative trigonometric shuffle algebra ${\mathrm A}$ is a space of symmetric rational functions satisfying certain wheel conditions. We describe a ring isomorphism between ${\mathrm A}$ and the center of the Hecke algebra using a realization of the elements of ${\mathrm A}$ as partition functions of coloured lattice paths associated to the $R$-matrix of…
▽ More
The commutative trigonometric shuffle algebra ${\mathrm A}$ is a space of symmetric rational functions satisfying certain wheel conditions. We describe a ring isomorphism between ${\mathrm A}$ and the center of the Hecke algebra using a realization of the elements of ${\mathrm A}$ as partition functions of coloured lattice paths associated to the $R$-matrix of $\mathcal U_{t^{1/2}}(\widehat{gl}_{\infty})$. As an application, we compute under certain conditions the Hilbert series of the commuting scheme and identify it with a particular element of the shuffle algebra ${\mathrm A}$, thus providing a combinatorial formula for it as a "domain wall" type partition function of coloured lattice paths.
△ Less
Submitted 14 February, 2022; v1 submitted 14 October, 2021;
originally announced October 2021.
-
The $R$-matrix of the quantum toroidal algebra $U_{q,t}(\overset{..}{gl}_1)$ in the Fock module
Authors:
Alexandr Garbali,
Jan de Gier
Abstract:
We propose a method to compute the $R$-matrix $R$ on a tensor product of Fock modules from coproduct relations in a Hopf algebra. We apply this method to the quantum toroidal algebra $U_{q,t}(\overset{..}{gl}_1)$ for which $R$ is currently not explicitly known. We show that the coproduct relations of $U_{q,t}(\overset{..}{gl}_1)$ reduce to a single elegant equation for $R$. Using the theory of sym…
▽ More
We propose a method to compute the $R$-matrix $R$ on a tensor product of Fock modules from coproduct relations in a Hopf algebra. We apply this method to the quantum toroidal algebra $U_{q,t}(\overset{..}{gl}_1)$ for which $R$ is currently not explicitly known. We show that the coproduct relations of $U_{q,t}(\overset{..}{gl}_1)$ reduce to a single elegant equation for $R$. Using the theory of symmetric Macdonald polynomials we show that this equation provides a recursive formula for the matrix elements of $R$.
△ Less
Submitted 13 May, 2020; v1 submitted 20 April, 2020;
originally announced April 2020.
-
Modified Macdonald polynomials and integrability
Authors:
Alexandr Garbali,
Michael Wheeler
Abstract:
We derive combinatorial formulae for the modified Macdonald polynomial $H_λ(x;q,t)$ using coloured paths on a square lattice with quasi-cylindrical boundary conditions. The derivation is based on an integrable model associated to the quantum group of $U_{q}(\wide{sl}_{n+1})$.
We derive combinatorial formulae for the modified Macdonald polynomial $H_λ(x;q,t)$ using coloured paths on a square lattice with quasi-cylindrical boundary conditions. The derivation is based on an integrable model associated to the quantum group of $U_{q}(\wide{sl}_{n+1})$.
△ Less
Submitted 12 November, 2019; v1 submitted 29 October, 2018;
originally announced October 2018.
-
A curious mapping between supersymmetric quantum chains
Authors:
Gyorgy Z. Feher,
Alexandr Garbali,
Jan de Gier,
Kareljan Schoutens
Abstract:
We present a unitary transformation relating two apparently different supersymmetric lattice models in one dimension. The first \cite{FS07} describes semionic particles on a 1D ladder, with supersymmetry moving particles between the two legs. The second \cite{GFNR15} is a fermionic model with particle-hole symmetry and with supersymmetry creating or annihilating pairs of domain walls. The mapping…
▽ More
We present a unitary transformation relating two apparently different supersymmetric lattice models in one dimension. The first \cite{FS07} describes semionic particles on a 1D ladder, with supersymmetry moving particles between the two legs. The second \cite{GFNR15} is a fermionic model with particle-hole symmetry and with supersymmetry creating or annihilating pairs of domain walls. The mapping we display features non-trivial phase factors that generalise the sign factors occurring in the Jordan-Wigner transformation. We dedicate this work to our friend and colleague Bernard Nienhuis, on the occasion of his 65-th birthday.
△ Less
Submitted 7 October, 2017;
originally announced October 2017.
-
Koornwinder polynomials and the stationary multi-species asymmetric exclusion process with open boundaries
Authors:
Luigi Cantini,
Alexandr Garbali,
Jan de Gier,
Michael Wheeler
Abstract:
We prove that the normalisation of the stationary state of the multi-species asymmetric simple exclusion process (mASEP) is a specialisation of a Koornwinder polynomial. As a corollary we obtain that the normalisation of mASEP factorises as a product over multiple copies of the two-species ASEP.
We prove that the normalisation of the stationary state of the multi-species asymmetric simple exclusion process (mASEP) is a specialisation of a Koornwinder polynomial. As a corollary we obtain that the normalisation of mASEP factorises as a product over multiple copies of the two-species ASEP.
△ Less
Submitted 30 June, 2016;
originally announced July 2016.
-
A new generalisation of Macdonald polynomials
Authors:
Alexandr Garbali,
Jan de Gier,
Michael Wheeler
Abstract:
We introduce a new family of symmetric multivariate polynomials, whose coefficients are meromorphic functions of two parameters $(q,t)$ and polynomial in a further two parameters $(u,v)$. We evaluate these polynomials explicitly as a matrix product. At $u=v=0$ they reduce to Macdonald polynomials, while at $q=0$, $u=v=s$ they recover a family of inhomogeneous symmetric functions originally introdu…
▽ More
We introduce a new family of symmetric multivariate polynomials, whose coefficients are meromorphic functions of two parameters $(q,t)$ and polynomial in a further two parameters $(u,v)$. We evaluate these polynomials explicitly as a matrix product. At $u=v=0$ they reduce to Macdonald polynomials, while at $q=0$, $u=v=s$ they recover a family of inhomogeneous symmetric functions originally introduced by Borodin.
△ Less
Submitted 23 May, 2016;
originally announced May 2016.
-
The dilute Temperley-Lieb O($n=1$) loop model on a semi infinite strip: the sum rule
Authors:
A. Garbali,
B. Nienhuis
Abstract:
This is the second part of our study of the ground state eigenvector of the transfer matrix of the dilute Temperley-Lieb loop model with the loop weight $n=1$ on a semi infinite strip of width $L$. We focus here on the computation of the normalization (otherwise called the sum rule) $Z_L$ of the ground state eigenvector, which is also the partition function of the critical site percolation model.…
▽ More
This is the second part of our study of the ground state eigenvector of the transfer matrix of the dilute Temperley-Lieb loop model with the loop weight $n=1$ on a semi infinite strip of width $L$. We focus here on the computation of the normalization (otherwise called the sum rule) $Z_L$ of the ground state eigenvector, which is also the partition function of the critical site percolation model. The normalization $Z_L$ is a symmetric polynomial in the inhomogeneities of the lattice $z_1,..,z_L$. This polynomial satisfies several recurrence relations which we solve independently in terms of Jacobi-Trudi like determinants. Thus we provide a few determinantal expressions for the normalization $Z_L$.
△ Less
Submitted 3 January, 2017; v1 submitted 26 November, 2014;
originally announced November 2014.
-
The dilute Temperley-Lieb O($n=1$) loop model on a semi infinite strip: the ground state
Authors:
A. Garbali,
B. Nienhuis
Abstract:
We consider the integrable dilute Temperley-Lieb (dTL) O($n=1$) loop model on a semi-infinite strip of finite width $L$. In the analogy with the Temperley-Lieb (TL) O($n=1$) loop model the ground state eigenvector of the transfer matrix is studied by means of a set of $q$-difference equations, sometimes called the $q$KZ equations. We compute some ground state components of the transfer matrix of t…
▽ More
We consider the integrable dilute Temperley-Lieb (dTL) O($n=1$) loop model on a semi-infinite strip of finite width $L$. In the analogy with the Temperley-Lieb (TL) O($n=1$) loop model the ground state eigenvector of the transfer matrix is studied by means of a set of $q$-difference equations, sometimes called the $q$KZ equations. We compute some ground state components of the transfer matrix of the dTL model, and show that all ground state components can be recovered for arbitrary $L$ using the $q$KZ equation and certain recurrence relation. The computations are done for generic open boundary conditions.
△ Less
Submitted 3 January, 2017; v1 submitted 25 November, 2014;
originally announced November 2014.
-
The scalar product of XXZ spin chain revisited. Application to the ground state at $Δ=-1/2$
Authors:
Alexander Garbali
Abstract:
For the scalar product $S_n$ of the XXZ $s=1/2$ spin chain we derive a new determinant expression which is symmetric in the Bethe roots. We consider an application of this formula to the inhomogeneous groundstate of the model with $Δ=-1/2$ with twisted periodic boundary conditions. At this point the ground state eigenvalue $τ_n$ of the transfer matrix is known and has a simple form that does not c…
▽ More
For the scalar product $S_n$ of the XXZ $s=1/2$ spin chain we derive a new determinant expression which is symmetric in the Bethe roots. We consider an application of this formula to the inhomogeneous groundstate of the model with $Δ=-1/2$ with twisted periodic boundary conditions. At this point the ground state eigenvalue $τ_n$ of the transfer matrix is known and has a simple form that does not contain the Bethe roots. We use the knowledge of $τ_n(μ)$ to obtain a closed expression for the scalar product. The result is written in terms of Schur functions. The computations of the normalization of the ground state and the expectation value of $σ^z$ are also presented.
△ Less
Submitted 11 November, 2014;
originally announced November 2014.
-
The domain wall partition function for the Izergin-Korepin 19-vertex model at a root of unity
Authors:
Alexander Garbali
Abstract:
We study the domain wall partition function $Z_N$ for the $U_q(A_2^{(2)})$ (Izergin-Korepin) integrable $19$-vertex model on a square lattice of size $N$. $Z_N$ is a symmetric function of two sets of parameters: horizontal $ζ_1,..,ζ_N$ and vertical $z_1,..,z_N$ rapidities. For generic values of the parameter $q$ we derive the recurrence relation for the domain wall partition function relating…
▽ More
We study the domain wall partition function $Z_N$ for the $U_q(A_2^{(2)})$ (Izergin-Korepin) integrable $19$-vertex model on a square lattice of size $N$. $Z_N$ is a symmetric function of two sets of parameters: horizontal $ζ_1,..,ζ_N$ and vertical $z_1,..,z_N$ rapidities. For generic values of the parameter $q$ we derive the recurrence relation for the domain wall partition function relating $Z_{N+1}$ to $P_N Z_N$, where $P_N$ is the proportionality factor in the recurrence, which is a polynomial symmetric in two sets of variables $ζ_1,..,ζ_N$ and $z_1,..,z_N$. After setting $q^3=-1$ the recurrence relation simplifies and we solve it in terms of a Jacobi-Trudi-like determinant of polynomials generated by $P_N$.
△ Less
Submitted 29 October, 2018; v1 submitted 11 November, 2014;
originally announced November 2014.
