The genus zero log/local/open correspondence for quasi-tame Looijenga pairs

A Looijenga pair Y(D)=(Y,D)Y(D)=(Y,D) consists of a smooth rational complex projective surface YY and an anticanonical simple normal crossings divisor D=i=1lDiD=\bigcup_{i=1}^lD_i, with each DiD_i smooth, irreducible, and nef. Write

EY(D)=Tot(iOY(Di)).E_{Y(D)}=\operatorname{Tot}\left(\bigoplus_i\mathcal{O}_Y(-D_i)\right).

A nef pair is quasi-tame if EY(D)E_{Y(D)} is deformation-equivalent to the total space associated to a tame Looijenga pair. Let Yop(D)Y^{\rm op}(D) denote the corresponding open geometry, let dH2(Y(D),Z)d\in H_2(Y(D),\mathbb{Z}) be a curve class, and let Nd(EY(D))N_d(E_{Y(D)}), O0,d(Yop(D))O_{0,d}(Y^{\rm op}(D)), and N0,dlog(Y(D))N_{0,d}^{\rm \log}(Y(D)) denote the genus-zero local, open, and logarithmic Gromov--Witten invariants. Let GVd\mathrm{GV}_d, LMOV0,d\mathrm{LMOV}_{0,d}, and DTdnum\mathrm{DT}^{\rm num}_d denote the associated BPS and numerical Donaldson--Thomas invariants, and let Q(Y(D))\mathsf{Q}(Y(D)) be the associated quiver.

The genus zero log/local/open correspondence. The invariants satisfy

Nd(EY(D))=O0,d(Yop(D))=(j=1l(1)dDj1dDj)N0,dlog(Y(D)),N_d(E_{Y(D)})=O_{0,d}(Y^{\rm op}(D))=\left(\prod_{j=1}^l\frac{(-1)^{d\cdot D_j-1}}{d\cdot D_j}\right)N_{0,d}^{\rm \log}(Y(D)),

and

GVd(EY(D))=LMOV0,d(Yop(D))Z.\mathrm{GV}_d(E_{Y(D)})=\mathrm{LMOV}_{0,d}(Y^{\rm op}(D))\in\mathbb{Z}.

Moreover, if l=2l=2,

DTdnum(Q(Y(D)))=GVd(EY(D)).\mathrm{DT}^{\rm num}_d(\mathsf{Q}(Y(D)))=|\mathrm{GV}_d(E_{Y(D)})|.

This correspondence is a proposed identification of genus-zero log, local, open, and, when l=2l=2, quiver Donaldson--Thomas theories; the source states that it was proved in the cited work of Bousseau for the cases under consideration.

Sources & referencesView supporting material

Primary source

Andrea Brini and Yannik Schuler, “On quasi-tame Looijenga pairs”, arXiv:2201.01645 (2023).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.