Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Affine Algebras, Langlands Duality and Bethe Ansatz

Edward Frenkel

TL;DR

The work surveys how affine algebras at the critical level give rise to a geometric Langlands framework via Beilinson–Drinfeld localization, with the center identified with a classical W-algebra and opers encoding central characters. It connects this representation-theoretic structure to integrable systems, notably the Gaudin model, where degenerations and Bethe ansatz solve for joint spectra and link to opers with trivial monodromy. In genus zero, Drinfeld–Beilinson constructions align with Sklyanin’s separation of variables, revealing a deep equivalence between two realizations of the Langlands correspondence and suggesting quantum (q-deformed) parallels. The discussion extends to Wakimoto modules, Miura transformations, Hecke operators, and global versus local Langlands phenomena, illustrating a cohesive picture intertwining D-modules, opers, Gaudin models, and geometric representation theory. The results illuminate how geometric Langlands interacts with integrable systems and hints at quantum Langlands via quantum affine algebras and q-difference equations.

Abstract

We review various aspects of representation theory of affine algebras at the critical level, geometric Langlands correspondence, and Bethe ansatz in the Gaudin models. Geometric Langlands correspondence relates D-modules on the moduli space of G-bundles on a complex curve X and flat G^L-bundles on X. Beilinson and Drinfeld construct it by applying a localization functor to representations of affine algebras of critical level. We show that in genus zero the corresponding D-modules are closely related to the diagonalization problem in the Gaudin model associated to G. This allows us to give a new interpretation of the Bethe ansatz and Sklyanin's separation of variables in the Gaudin model in terms of Langlands correspondence.

Affine Algebras, Langlands Duality and Bethe Ansatz

TL;DR

The work surveys how affine algebras at the critical level give rise to a geometric Langlands framework via Beilinson–Drinfeld localization, with the center identified with a classical W-algebra and opers encoding central characters. It connects this representation-theoretic structure to integrable systems, notably the Gaudin model, where degenerations and Bethe ansatz solve for joint spectra and link to opers with trivial monodromy. In genus zero, Drinfeld–Beilinson constructions align with Sklyanin’s separation of variables, revealing a deep equivalence between two realizations of the Langlands correspondence and suggesting quantum (q-deformed) parallels. The discussion extends to Wakimoto modules, Miura transformations, Hecke operators, and global versus local Langlands phenomena, illustrating a cohesive picture intertwining D-modules, opers, Gaudin models, and geometric representation theory. The results illuminate how geometric Langlands interacts with integrable systems and hints at quantum Langlands via quantum affine algebras and q-difference equations.

Abstract

We review various aspects of representation theory of affine algebras at the critical level, geometric Langlands correspondence, and Bethe ansatz in the Gaudin models. Geometric Langlands correspondence relates D-modules on the moduli space of G-bundles on a complex curve X and flat G^L-bundles on X. Beilinson and Drinfeld construct it by applying a localization functor to representations of affine algebras of critical level. We show that in genus zero the corresponding D-modules are closely related to the diagonalization problem in the Gaudin model associated to G. This allows us to give a new interpretation of the Bethe ansatz and Sklyanin's separation of variables in the Gaudin model in terms of Langlands correspondence.

Paper Structure

This paper contains 40 sections, 6 theorems, 49 equations.

Table of Contents

  1. Introduction.
  2. Localization functor.
  3. Setup.
  4. Moduli space.
  5. Representations of critical level.
  6. Null-vectors and central elements.
  7. Symbols of null-vectors.
  8. The structure of the center.
  9. The structure of the center.
  10. Hamiltonian reduction.
  11. Opers.
  12. $\widehat{{\mathfrak g}}$--modules associated to opers.
  13. Wakimoto modules.
  14. Miura transformation.
  15. $\widehat{{\mathfrak g}}$--modules and connections.
  16. ...and 25 more sections

Key Result

Theorem 1

There exist $S^1,\ldots,S^l \in {\mathfrak z}(\widehat{{\mathfrak g}})$, such that $\deg S^i = d_i+1$, and ${\mathfrak z}(\widehat{{\mathfrak g}}) \simeq {\mathbb C}[\partial_t^n S^i]_{i=1,\ldots,l;n\geq 0}$. In particular, $S^1 = \frac{1}{2} \sum_a (J_a \otimes t^{-1})^2 v$.

Theorems & Definitions (7)

  • Theorem 1: FF:ctr
  • Theorem 2: FF:ctr
  • Proposition 1
  • Theorem 3: A. Beilinson, V. Drinfeld BD:quant
  • Conjecture 1
  • Proposition 2
  • Proposition 3