Do Langlands duality predictions match explicit computations for higher-rank gauge groups yet?

# Langlands Duality and Higher-Rank Gauge Groups: Recent Progress

Your question touches one of the most ambitious meeting points between pure mathematics and physics. The short answer is: yes, but with important caveats about what "explicit computation" means.

What has been confirmed

The preprints show concrete matching in several regimes:

For finite subgroups and specific pairs: The physics prediction that the number of homomorphisms from finite subgroups of SU(2) to a group G equals the number to its Langlands dual $\tilde{G}$ has been proven for important cases [AG-2025.05-883]. This includes the pairs $(SU(n), PU(n))$ and $(Sp(n), SO(2n+1))$ "for arbitrary Γ" — meaning the duality holds not just for cyclic subgroups, but for general non-Abelian finite groups. The authors explicitly note this is new territory and ask for a unified proof.

For multiplicative Hitchin systems: Higher-rank gauge groups now have a rigorous geometric matching. Twisted and untwisted multiplicative Hitchin fibrations associated with a simply-laced group $G$ have been shown to be "dual" to those of the Langlands dual $G^\vee$ in a precise sense: their bases coincide and their generic fibres are dual Beilinson motives [AG-2025.09-939]. This extends beyond the classical case and handles diagram automorphisms — a source of subtlety in higher-rank theories.

For analytic spectral problems: The analytic Langlands correspondence (solving spectral problems for quantized Hitchin Hamiltonians) has been extended to higher rank via a beautiful route: the Riemann-Hilbert problem for arbitrary-rank oper connections now admits explicit solutions in terms of a nonlinear integral equation, and the generating function matches the Yang-Yang function of the quantum Toda chain [AG-2025.12-1157]. This proves a conjecture of Nekrasov, Rosly and Shatashvili — a concrete computational victory.

What remains open

The landscape is not yet fully unified:

• The authors of [AG-2025.05-883] explicitly ask for a uniform proof covering all cases simultaneously, suggesting case-by-case verification is still the norm.
• Extensions to non-simply-laced groups are conjectured [AG-2025.10-1020] but not yet fully established in all settings.
• The connection between different formulations (categorical Floer homologies, quantum K-theory, spectral curves) is being built but remains active research [AG-2025.11-937], [AG-2026.04-800].

Why this matters

These results aren't academic trivia. Langlands duality predicts hidden symmetries in gauge theories — swap the gauge group for its dual and you get a seemingly different theory that is secretly the same [AG-2024.02-510]. Verifying this for higher ranks means the symmetry survives the jump from rank-1 (abelian) to rank-$n$ non-commutative gauge groups, making it robust enough to guide physics and inform new mathematics.