Is bipartite Quantum Max-Cut in BPP?
- Field
- Topics
Problem
Is the following bipartite Quantum Max-Cut promise problem in \(\mathrm{BPP}\)? Given a bipartite graph \(G=(V,E)\) with \(|V|=n\), polynomially bounded nonnegative rational weights \(w_{ij}\), and rational thresholds \(a<b\) with \(b-a\geq1/\operatorname{poly}(n)\), define
Here \(X_i,Y_i,Z_i\) are Pauli operators on qubit \(i\). For the energy in Eq. (1), distinguish \(E_0\leq a\) from \(E_0\geq b\), promised one holds, using a randomized classical algorithm polynomial in the input length and correct with probability at least \(2/3\).
Source
The remaining classical part of Open question 3.4 in Gharibian’s The 7 faces of quantum NP, Section 3, arXiv page 7 [Gha24]. The original question asks for the complexity of bipartite QMC; the present formulation focuses on \(\mathrm{BPP}\) membership following the 2026 \(\mathrm{BQP}\) upper bound [RT26].
Progress
Earlier upper bound: conjugating one bipartition by Pauli \(Y\) makes \(H_G\) stoquastic, placing bipartite QMC in \(\mathrm{StoqMA}\) (summarized in Section 3) [Gha24].
July 2026: Rayudu–Takahashi prove a Lee-Yang spectral gap at least \(h/4\) under field strength \(h>0\), enabling adiabatic ground-energy estimation to inverse-polynomial additive error and establishing \(\mathrm{BQP}\) membership (Theorem 11; Section 1.1) [RT26].
Independently, Bravyi–Gosset–Liu–Wong give a \(\operatorname{poly}(n,J,1/\epsilon)\)-time quantum algorithm for additive-\(\epsilon\) ground-energy estimation on weighted bipartite graphs, where \(J=\max_{\{i,j\}\in E}w_{ij}\) (Corollary 2) [BGLW26].
Comment
Membership in \(\mathrm{BPP}\) remains open: the quantum upper bound leaves classical complexity unresolved [RT26]. Here \(\mathrm{BPP}\) and \(\mathrm{BQP}\) denote their promise-problem versions. The target is inverse-polynomial additive precision.