DT/PT correspondence conjecture with descendents and Euler-class corrections

Let XX be a nonsingular threefold, let β\beta be a curve class, let αiH>0(X)\alpha_i\in H^{>0}(X), and let ki,ljk_i,l_j be positive integers. Set e=c1(X)c2(X)c3(X)\mathrm{e}=c_1(X)c_2(X)-c_3(X) and define

M(q)=i=1(1qi)i,F1=qddqlogM(q),M(q)=\prod_{i=1}^{\infty}(1-q^i)^{-i},\qquad \mathfrak{F}_1=q\frac{d}{dq}\log M(-q), F1(i)=(qddq)iF1.\mathfrak{F}_1^{(i)}=\left(q\frac{d}{dq}\right)^i\mathfrak{F}_1.

For disjoint subsets Jr{1,,m}J_r\subset\{1,\dots,m\}, put J={1,,m}rJr\overline{J}=\{1,\dots,m\}\setminus\bigcup_rJ_r, Jr=jJr(lj3)\lVert J_r\rVert=\sum_{j\in J_r}(l_j-3), f(J)=r=1sf(Jr)f(J)=\prod_{r=1}^sf(J_r), and

f(Jr)=(1)Jr1F1(Jr1)Jr!jJr(lj3)!.f(J_r)=(-1)^{|J_r|-1}\mathfrak{F}_1^{(|J_r|-1)}\frac{\lVert J_r\rVert!}{\prod_{j\in J_r}(l_j-3)!}.

DT/PT correspondence conjecture. For any such data,

ZDT,β,X(i=1nchki(αi)j=1mchlj(1))=ZPT,βX(ichki(αi)jchlj(1))+s=1mJ1,,Jsf(J)ZPT,βX(i=1nchki(αi)r=1schJr(e)jJchlj(1)).\begin{aligned} \mathrm{Z}^{\prime,X}_{\rm DT,\beta}\left(\prod_{i=1}^n\mathrm{ch}_{k_i}(\alpha_i)\prod_{j=1}^m\mathrm{ch}_{l_j}(1)\right) &=\mathrm{Z}^{X}_{\rm PT,\beta}\left(\prod_i\mathrm{ch}_{k_i}(\alpha_i)\prod_j\mathrm{ch}_{l_j}(1)\right)\\ &\quad+\sum_{s=1}^m\sum_{J_1,\dots,J_s}f(J)\,\mathrm{Z}^{X}_{\rm PT,\beta}\left(\prod_{i=1}^n\mathrm{ch}_{k_i}(\alpha_i)\prod_{r=1}^s\mathrm{ch}_{\lVert J_r\rVert}(\mathrm{e})\prod_{j\in\overline{J}}\mathrm{ch}_{l_j}(1)\right). \end{aligned}

The formula gives a regularized expression for DT descendents in terms of PT invariants and universal degree-zero correction functions.

Sources & referencesView supporting material

Primary source

A. Oblomkov, “EGL formula for DT/PT theory of local curves”, arXiv:1901.03014 (2019).

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.