Abstract
We study 3d $\mathcal{N}=2$ supersymmetric gauge theories on the product of a line and a Riemann surface. By performing a topological twist along the Riemann surface, we derive an effective supersymmetric quantum mechanics on the line. We propose a construction of the space of supersymmetric ground states—the "twisted Hilbert space"—as a graded vector space in terms of the cohomology of generalized vortex moduli spaces. This framework allows for a more sensitive probe of 3d abelian mirror symmetry than traditional numerical indices, accounting for superpotentials, real masses, and background flavor bundles.
Introduction: Beyond Numerical Indices
In the study of supersymmetric (SUSY) field theories, "indices" are standard tools used to count protected states. The supersymmetric twisted index is a powerful invariant for 3d theories, but it is fundamentally a "coarse" measurement. As a graded trace (a weighted sum of states with $\pm 1$ signs), it often hides the true complexity of a theory through massive cancellations.
To gain a deeper understanding, we must move beyond counting and describe the Hilbert space itself—the actual vector space of ground states. This is particularly crucial for testing dualities like 3D mirror symmetry, where two seemingly different theories are conjectured to be physically identical. A robust duality should match the entire Hilbert space, not just a single numerical representative.
Theoretical Framework: Topological Twisting and Reduction
The core of this approach involves placing a 3d $\mathcal{N}=2$ theory on a specific manifold: $\mathbb{R} \times C$, where $C$ is a Riemann surface of genus $g$.
By performing a topological twist along the Riemann surface $C$, we preserve a specific set of supercharges. This process effectively reduces the 3d physics to a 1d $\mathcal{N}=(0,2)$ Supersymmetric Quantum Mechanics (SQM) residing on the line $\mathbb{R}$. The ground states of this 1D system correspond to the supersymmetric states of the original 3d theory.
Geometric Realization via Vortex Moduli Spaces
The authors identify these ground states with the $L^2$-cohomology of a target space. In the "Higgs branch" regime, this target space is the moduli space of generalized vortices on the Riemann surface $C$.
Key aspects of this construction include:
- Vortices as Ground States: SUSY ground states are encoded in the geometry of vortices, specifically as sections of a $\mathbb{Z}_2$-graded vector bundle over the vortex moduli space.
- Chamber Physics: The Hilbert space is piecewise constant. As real mass parameters vary, the space can "jump" at specific hyperplanes where new massless degrees of freedom emerge.
- Deformation Sensitivity: Unlike the twisted index, the twisted Hilbert space is sensitive to 3d superpotential deformations, making it a "higher-fidelity" tool for duality verification.
Key Results and Duality Verification
The paper provides explicit graded vector space descriptions for several foundational models, including single chiral multiplets, the XYZ model (three chirals with a cubic superpotential), and U(1) SQED (two chirals with opposite charges).
A major success of this work is the verification of 3d abelian mirror symmetry. The authors demonstrate that for dual pairs like the XYZ model and U(1) SQED, the complex cohomology of their respective (and geometrically distinct) vortex moduli spaces matches perfectly: $\mathcal{H}_{XYZ} \cong \mathcal{H}_{SQED}$. This holds even for Riemann surfaces with genus $g > 0$, where previous numerical checks were computationally prohibitive.
Conclusion and Future Directions
This work establishes the twisted Hilbert space as a rigorous and refined invariant for 3d SUSY theories. While currently focused on abelian gauge groups, future work aims to extend this to non-abelian theories ($G=SU(N)$), which involve more complex "quotient stack" moduli spaces. Additionally, the framework could be expanded to include SUSY line operators, potentially revealing deeper links to the 3d-3d correspondence and 4d Donaldson-Witten theory.
Summary updated