DT4/Gopakumar–Vafa correspondence for holomorphic symplectic 4-folds

About 4 years old · traced to

Let XX be a holomorphic symplectic 44-fold and let MβM_\beta be the moduli scheme of one-dimensional stable sheaves FF on XX with ch⁡3(F)=β\operatorname{ch}_3(F)=\beta and χ(F)=1\chi(F)=1. Choose a reduced virtual class [Mβ]vir∈A2(Mβ,Q)[M_\beta]^{\mathrm{vir}}\in A_2(M_\beta,\mathbb{Q}) and an orientation, and define DT⁡4\operatorname{DT}_4 descendent invariants by

⟨τk1(γ1),…,τkn(γn)⟩βDT⁡4=∫[Mβ]vir∏i=1nτki(γi).\left\langle\tau_{k_1}(\gamma_1),\ldots,\tau_{k_n}(\gamma_n)\right\rangle^{\operatorname{DT}_4}_\beta=\int_{[M_\beta]^{\mathrm{vir}}}\prod_{i=1}^n\tau_{k_i}(\gamma_i).

Let n0,β(γ1,…,γn)n_{0,\beta}(\gamma_1,\ldots,\gamma_n), n1,β(γ)n_{1,\beta}(\gamma), and n2,βn_{2,\beta} denote the Gopakumar–Vafa invariants defined in the paper. DT4/Gopakumar–Vafa correspondence. For a certain choice of orientation, if β\beta is an effective curve class, then

⟨τ0(γ1),…,τ0(γn)⟩βDT⁡4=n0,β(γ1,…,γn).\left\langle\tau_0(\gamma_1),\ldots,\tau_0(\gamma_n)\right\rangle^{\operatorname{DT}_4}_\beta=n_{0,\beta}(\gamma_1,\ldots,\gamma_n).

If β\beta is a primitive curve class, then

⟨τ1(γ)⟩βDT⁡4=−12⟨τ1(γ)⟩0,βGW⁡−n1,β(γ),\left\langle\tau_1(\gamma)\right\rangle^{\operatorname{DT}_4}_\beta=-\frac{1}{2}\left\langle\tau_1(\gamma)\right\rangle^{\operatorname{GW}}_{0,\beta}-n_{1,\beta}(\gamma),

and

−⟨τ3(1)⟩βDT⁡4−112⟨τ1(c2(X))⟩βDT⁡4=n2,β.-\left\langle\tau_3(1)\right\rangle^{\operatorname{DT}_4}_\beta-\frac{1}{12}\left\langle\tau_1(c_2(X))\right\rangle^{\operatorname{DT}_4}_\beta=n_{2,\beta}.

The conjecture proposes a sheaf-theoretic interpretation of all genus Gopakumar–Vafa invariants through reduced 44-dimensional Donaldson–Thomas invariants. The source gives no resolution status.

References

Primary source

Yalong Cao, Georg Oberdieck and Yukinobu Toda, “Gopakumar-Vafa type invariants of holomorphic symplectic 4-folds”, arXiv:2201.10878 (2022).

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.