Can topological string spectral dualities predict new quantum mirror curve invariants?

# Topological String Spectral Dualities and Quantum Mirror Curve Invariants

Your question touches on a rich intersection of topological string theory, quantum curves, and spectral dualities—all active areas in the archive. The short answer is: yes, there is compelling recent work suggesting these connections, though the field is still developing the precise predictive framework.

The Core Connection

The key insight is that spectral dualities in topological string theory give us multiple ways to encode the *same* geometric information. When you quantize the mirror curve of a toric Calabi-Yau threefold, you get a quantum operator whose spectral properties encode open and closed BPS invariants [AG-2026.01-761]. The duality aspect means that different "dual" perspectives on the same geometry—say, via $x$-$y$ duality or symplectic duality—should predict the same invariants but from different angles [AG-2025.04-1162].

This matters because it constrains which invariants can actually appear: if two dual descriptions must agree, you get non-trivial constraints on the quantum curve operators and the BPS numbers they count.

How Dualities Predict New Invariants

Several mechanisms are at work:

1. Topological Recursion via Duality

By leveraging $x$-$y$ duality and symplectic duality, researchers can now compute quantum spectral curve operators more efficiently [AG-2025.04-1162]. This extended topological recursion machinery—including recent logarithmic extensions [AG-2025.10-794]—allows you to bootstrap invariants from a minimal set of initial data. The dualities essentially give you more equations to solve, narrowing down which solutions are consistent.

2. Open/Closed Duality and Wall-Crossing

For strip geometries (a subclass of toric Calabi-Yau threefolds), open topological string partition functions generate both open Gromov-Witten and Donaldson-Thomas invariants via topological recursion [AG-2025.10-794]. The associated quantum curve encodes how 3D-5D wall crossing affects these invariants as parameters vary. The duality here predicts relationships between open invariants that weren't obvious before.

3. Langlands Duality of Categories

At a more abstract level, mirror symmetry in 5d topological gauge theories exhibits Langlands duality of $A_\infty$-categories of Floer homologies [AG-2025.11-937]. This suggests that dual gauge groups predict dual categories of invariants—a very general predictive principle.

Concrete Predictions

The most concrete results so far come from specific examples:

• For local $\mathbb{P}^2$, the quantum curve's fermionic spectral trace exhibits a full-fledged strong-weak resurgent symmetry, where perturbative and non-perturbative contributions exchange roles under duality [AG-2024.04-779]. This constrains the structure of BPS invariants across coupling regimes.

• For Aganagic-Vafa branes on special Lagrangian submanifolds, localization calculations produce generating functions of refined open BPS invariants in a polarization determined by the brane data [AG-2026.01-761]. The quantum curve quantization (in the Nekrasov-Shatashvili limit) is unambiguous—the brane geometry picks out a unique quantization, predicting specific invariants.

• The skein-valued mirror construction shows that holomorphic curve counts with boundaries on three special Lagrangian tori satisfy skein-valued operator equations that dequantize to a mirror geometry [AG-2024.12-1001]. The algebra uniquely determines the count, demonstrating predictive power.

The Frontier Question

What's not yet fully settled is whether you can *a priori* list all quantum mirror curve invariants just from the spectral duality structure, without computing them explicitly. The archive suggests you can make strong predictions—ruling out certain invariants, constraining relationships between them—but the full "dictionary" is still being written. Most results so far are either derived from explicit recursion (which requires some initial data) or proven post-hoc for specific geometries.