Comments on F-maximization and R-symmetry in 3D SCFTs
Vasilis Niarchos
TL;DR
This work tests the $F$-maximization principle as a nonperturbative method to determine the exact $U(1)$ R-symmetry in 3D ${\cal N}=2$ SCFTs by analyzing a large-$N$ Chern-Simons-Matter theory with a single adjoint. Using localization, the $S^3$ partition function is reduced to a matrix model and solved numerically in the saddle-point regime to extract $R(\lambda)$ as a function of the 't Hooft coupling $\lambda=N/k$, with $F=-\log|Z_{S^3}|$ maximized over trial R-charges. The results show $R(\lambda)$ monotonically decreases and remains near the nonperturbative bounds, matching perturbative expectations at small $\lambda$, but reveal tensions with a proposed Seiberg-like duality, indicating possible necessary refinements of the $F$-maximization framework or decoupling-field effects. The paper further conjectures a qualitative picture for the exact R-symmetry of the ${\bf \hat{A}}$ theory at strong coupling, including potential oscillatory behavior around a simple bounding curve and implications for deformations $W_{n+1}=\mathrm{Tr} X^{n+1}$ and for the $F$-theorem.
Abstract
We report preliminary results on the recently proposed F-maximization principle in 3D SCFTs. We compute numerically in the large-N limit the free energy on the three-sphere of an N=2 Chern-Simons-Matter theory with a single adjoint chiral superfield which is known to exhibit a pattern of accidental symmetries associated to chiral superfields that hit the unitarity bound and become free. We observe that the F-maximization principle produces a U(1) R-symmetry consistent with previously obtained bounds but inconsistent with a postulated Seiberg-like duality. Potential modifications of the principle associated to the decoupling fields do not appear to be sufficient to account for the observed violations.
