The commuting local-Hamiltonian (CLH) problem is a restriction of the local-Hamiltonian problem, in which the terms of the Hamiltonian are required to pairwise commute. A long line of work has shown that the problem lies in $\mathsf{NP}$ for certain families of commuting local Hamiltonians. Nevertheless, there has been no formal evidence against the possibility that the general CLH problem is $\mathsf{QMA}$-complete.
CLH is complete for the complexity class $\mathsf{QIMA}$, defined through quantum verifiers whose local gates commute. Therefore, CLH is $\mathsf{QMA}$-complete if and only if $\mathsf{QIMA}=\mathsf{QMA}$. In this work, we introduce a classical-oracle analogue $\mathsf{QIMA}^{\mathcal{O}}$ and construct a classical oracle $\mathcal{O}$ such that $\mathsf{BQP}^{\mathcal{O}}\not\subseteq\mathsf{QIMA}^{\mathcal{O}}$.
Since $\mathsf{BQP}^{\mathcal{O}}\subseteq\mathsf{QMA}^{\mathcal{O}}$ for any classical oracle $\mathcal{O}$, this implies $\mathsf{QIMA}^{\mathcal{O}}\neq\mathsf{QMA}^{\mathcal{O}}$ for our constructed oracle. Thus, our result provides relativized evidence against the possibility that the general commuting local-Hamiltonian problem is $\mathsf{BQP}$-hard, and hence also against the possibility that it is $\mathsf{QMA}$-complete.
AI Summary
The paper asks whether the general commuting local-Hamiltonian (CLH) problem could be BQP-hard or QMA-complete. Using its characterization by the commuting-verifier class $\mathsf{QIMA}$, it defines a classical-oracle extension and constructs an oracle $O$ such that $\mathsf{BQP}^O\not\subseteq\mathsf{QIMA}^O$, hence $\mathsf{QIMA}^O\ne\mathsf{QMA}^O$.
The oracle model permits polynomial-size, nonlocal commuting units, tested through Hadamard tests on an untrusted witness. Oracle-containing units must be exact reflections; oracle-free units may be arbitrary unitaries. There is no trusted workspace beyond the test controls. For Forrelation on $N=2^n$ truth-table entries, any such verifier requires $C(n)+T(n)\ge\beta2^n-O(1)$ for even $n$, where $C,T$ count classical preprocessing and quantum queries, and $\beta$ is the NO-promise threshold. BQP uses constantly many queries. The proof freezes a preprocessing transcript, expresses the dimension of a high-acceptance subspace as a degree-$T$ polynomial, and uses bent-function instances, Hamming-layer symmetrization, and root counting. Diagonalization yields the separating oracle.
The separation extends to specified negligible-error implementations and additional non-reflection one-query units whose surrounding circuits touch at most $k$ address qubits, giving $C+2^kT\ge\beta2^n-O(1)$. Conversely, allowing inverse-polynomial deviations from reflections, unrestricted non-reflection one-query units, or one additional trusted active qubit recovers $\mathsf{QMA}^O$.
The conclusion is relativized evidence, not an unrelativized hardness exclusion. Relativization is noncanonical: the chosen model drops locality and equals QMA for the identity oracle. Negligible-error robustness requires uniformly supplied, commuting, query-preserving exact companions; whether perfect completeness changes the oracle class is unresolved.
Likely largely AI-generatedAI use disclosedStrong assumptions
ChatGPT 6 Astra was then used for proving sections \ref{subsec:restricted-address-support-lower-bound} and \ref{subsec:one-query-collapse}, and for improving the writing throughout the paper.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
The commuting local-Hamiltonian (CLH) problem is a restriction of the local-Hamiltonian problem, in which the terms of the Hamiltonian are required to pairwise commute. A long line of work has shown that the problem lies in NP for certain families of commuting local Hamiltonians. Nevertheless, there has been no formal evidence against the possibility that the general CLH problem is QMA-complete.
CLH is complete for the complexity class QIMA, defined through quantum verifiers whose local gates commute. Therefore, CLH is QMA-complete if and only if QIMA=QMA. In this work, we introduce a classical-oracle analogue QIMAO and construct a classical oracle O such that BQPO⊆QIMAO.
Since BQPO⊆QMAO for any classical oracle O, this implies QIMAO=QMAO for our constructed oracle. Thus, our result provides relativized evidence against the possibility that the general commuting local-Hamiltonian problem is BQP-hard, and hence also against the possibility that it is QMA-complete.
The paper asks whether the general commuting local-Hamiltonian (CLH) problem could be BQP-hard or QMA-complete. Using its characterization by the commuting-verifier class QIMA, it defines a classical-oracle extension and constructs an oracle O such that BQPO⊆QIMAO, hence QIMAO=QMAO.
The oracle model permits polynomial-size, nonlocal commuting units, tested through Hadamard tests on an untrusted witness. Oracle-containing units must be exact reflections; oracle-free units may be arbitrary unitaries. There is no trusted workspace beyond the test controls. For Forrelation on N=2n truth-table entries, any such verifier requires C(n)+T(n)≥β2n−O(1) for even n, where C,T count classical preprocessing and quantum queries, and β is the NO-promise threshold. BQP uses constantly many queries. The proof freezes a preprocessing transcript, expresses the dimension of a high-acceptance subspace as a degree-T polynomial, and uses bent-function instances, Hamming-layer symmetrization, and root counting. Diagonalization yields the separating oracle.
The separation extends to specified negligible-error implementations and additional non-reflection one-query units whose surrounding circuits touch at most k address qubits, giving C+2kT≥β2n−O(1). Conversely, allowing inverse-polynomial deviations from reflections, unrestricted non-reflection one-query units, or one additional trusted active qubit recovers QMAO.
The conclusion is relativized evidence, not an unrelativized hardness exclusion. Relativization is noncanonical: the chosen model drops locality and equals QMA for the identity oracle. Negligible-error robustness requires uniformly supplied, commuting, query-preserving exact companions; whether perfect completeness changes the oracle class is unresolved.
The local Hamiltonian problem is famously complete for the class QMA, the quantum analogue of NP. The complexity of its semi-classical version, in which the terms of the Hamiltonian are required to commute (the CLH problem), has attracted considerable attention recently due to its intriguing nature, as well as in relation to growing interest in the qPCP conjecture. We show here that if the underlying bipartite interaction graph of the CLH instance is a good locally-expanding graph, namely, the expansion of any constant-size set is e-close to optimal, then approximating its ground energy to within additive factor O(e) lies in NP. The proof holds for k- local Hamiltonians for any constant k and any constant dimensionality of particles d. We also show that the approximation problem of CLH on such good local expanders is NP-hard. This implies that too good local expansion of the interaction graph constitutes an obstacle against quantum hardness of the approximation problem, though it retains its classical hardness. The result highlights new difficulties in trying to mimic classical proofs (in particular Dinur's PCP proof) in an attempt to prove the quantum PCP conjecture. A related result was discovered recently independently by Brandao and Harrow, for 2-local general Hamiltonians, bounding the quantum hardness of the approximation problem on good expanders, though no NP-hardness is known in that case.
The complexity of the commuting local Hamiltonians (CLH) problem still remains a mystery after two decades of research of quantum Hamiltonian complexity; it is only known to be contained in NP for few low parameters. Of particular interest is the tightly related question of understanding whether groundstates of CLHs can be generated by efficient quantum circuits. The two problems touch upon conceptual, physical and computational questions, including the centrality of non-commutation in quantum mechanics, quantum PCP and the area law. It is natural to try to address first the more physical case of CLHs embedded on a 2D lattice but this problem too remained open, apart from some very specific cases. Here we consider a wide class of two dimensional CLH instances; these are k-local CLHs, for any constant k; they are defined on qubits set on the edges of any surface complex, where we require that this surface complex is not too far from being "Euclidean". Each vertex and each face can be associated with an arbitrary term (as long as the terms commute). We show that this class is in NP, and moreover that the groundstates have an efficient quantum circuit that prepares them. This result subsumes that of Schuch [2011] which regarded the special case of 4-local Hamiltonians on a grid with qubits, and by that it removes the mysterious feature of Schuch's proof which showed containment in NP without providing a quantum circuit for the groundstate and considerably generalizes it. We believe this work and the tools we develop make a significant step towards showing that 2D CLHs are in NP.
Commuting local Hamiltonians provide a testing ground for studying many of the most interesting open questions in quantum information theory, including the quantum PCP conjecture and the nature of entanglement. However, unlike the general local Hamiltonian problem, the exact complexity of the commuting local Hamiltonian problem (CLH) remains unknown. A number of works have shown that increasingly expressive families of commuting local Hamiltonians admit classical verifiers. Despite intense work, proofs placing CLH in NP rely heavily on an underlying 2D lattice structure, or a very constrained local dimension and locality. In this work, we present a new technique to analyze the complexity of various families of commuting local Hamiltonians: guided reductions. Intuitively, these are a generalization of typical reduction where the prover provides a guide so that the verifier can construct a simpler Hamiltonian. The core of our reduction is a new rounding technique based on a combination of Jordan’s Lemma for pairs of projectors and the Structure Lemma for C^* algebras. Our rounding technique is much more flexible than previous work and allows us to remove constraints on local dimension in exchange for a rank-1 assumption. Using our rounding technique, we prove the following two results: 1) 2D-CLH for rank-1 instances are contained in NP, independent of the qudit dimension. It is notable that this family of commuting local Hamiltonians has no restriction on the local dimension or the locality of the Hamiltonian terms. 2) 3D-CLH for rank-1 instances are in NP. To our knowledge this is the first time a family of {3D} commuting local Hamiltonians has been contained in NP. Our results apply to Hamiltonians with large qudit degree and remain non-trivial despite the quantum Lovász Local Lemma. [Andris Ambainis et al., 2012]
In this work we consider the ground space connectivity problem for commuting local Hamiltonians. The ground space connectivity problem asks whether it is possible to go from one (efficiently preparable) state to another by applying a polynomial length sequence of 2-qubit unitaries while remaining at all times in a state with low energy for a given Hamiltonian H . It was shown in [Gharibian and Sikora, ICALP15] that this problem is QCMA-complete for general local Hamiltonians, where QCMA is defined as QMA with a classical witness and BQP verifier. Here we show that the commuting version of the problem is also QCMA-complete. This provides one of the first examples where commuting local Hamiltonians exhibit complexity theoretic hardness equivalent to general local Hamiltonians.
Commuting local Hamiltonians provide a testing ground for studying many of the most interesting open questions in quantum information theory, including the quantum PCP conjecture and the existence of area laws. Although they are a simplified model of quantum computation, the status of the commuting local Hamiltonian problem remains largely unknown. A number of works have shown that increasingly expressive families of commuting local Hamiltonians admit completely classical verifiers. Despite intense work, the largest class of commuting local Hamiltonians we can place in NP are those on a square lattice, where each lattice site is a qutrit. Even worse, many of the techniques used to analyze these problems rely heavily on the geometry of the square lattice and the properties of the numbers 2 and 3 as local dimensions. In this work, we present a new technique to analyze the complexity of various families of commuting local Hamiltonians: guided reductions. Intuitively, these are a generalization of typical reduction where the prover provides a guide so that the verifier can construct a simpler Hamiltonian. The core of our reduction is a new rounding technique based on a combination of Jordan's Lemma and the Structure Lemma. Our rounding technique is much more flexible than previous work, and allows us to show that a larger family of commuting local Hamiltonians is in NP, albiet with the restriction that all terms are rank-1. Specifically, we prove the following two results: 1. Commuting local Hamiltonians in 2D that are rank-1 are contained in NP, independent of the qudit dimension. Note that this family of commuting local Hamiltonians has no restriction on the local dimension or the locality. 2. We prove that rank-1, 3D commuting Hamiltonians with qudits on edges are in NP. To our knowledge this is the first time a family of 3D commuting local Hamiltonians has been contained in NP.
The canonical hard problems for NP and its quantum analogue, Quantum Merlin-Arthur (QMA), are MAX-k-SAT and the k-local Hamiltonian problem (k-LH), the quantum generalization of MAX-k-SAT, respectively. In recent years, however, an arguably even more physically motivated problem than k-LH has been formalized - the problem of simulating local measurements on ground states of local Hamiltonians (APX-SIM). Perhaps surprisingly, [Ambainis, CCC 2014] showed that APX-SIM is likely harder than QMA. Indeed, [Ambainis, CCC 2014] showed that APX-SIM is P^{QMA[log]}-complete, for P^{QMA[log]} the class of languages decidable by a P machine making a logarithmic number of adaptive queries to a QMA oracle. In this work, we show that APX-SIM is P^{QMA[log]}-complete even when restricted to physically motivated Hamiltonians, obtaining as intermediate steps a variety of related complexity-theoretic results. Specifically, we first give a sequence of results which together yield P^{QMA[log]}-hardness for APX-SIM on well-motivated Hamiltonians such as the 2D Heisenberg model: - We show that for NP, StoqMA, and QMA oracles, a logarithmic number of adaptive queries is equivalent to polynomially many parallel queries. Formally, P^{NP[log]}=P^{||NP}, P^{StoqMA[log]}=P^{||StoqMA}, and P^{QMA[log]}=P^{||QMA}. (The result for NP was previously shown using a different proof technique.) These equalities simplify the proofs of our subsequent results. - Next, we show that the hardness of APX-SIM is preserved under Hamiltonian simulations (à la [Cubitt, Montanaro, Piddock, 2017]) by studying a seemingly weaker problem, ∀-APX-SIM. As a byproduct, we obtain a full complexity classification of APX-SIM, showing it is complete for P, P^{||NP},P^{||StoqMA}, or P^{||QMA} depending on the Hamiltonians employed. - Leveraging the above, we show that APX-SIM is P^{QMA[log]}-complete for any family of Hamiltonians which can efficiently simulate spatially sparse Hamiltonians. This implies APX-SIM is P^{QMA[log]}-complete even on physically motivated models such as the 2D Heisenberg model. Our second focus considers 1D systems: We show that APX-SIM remains P^{QMA[log]}-complete even for local Hamiltonians on a 1D line of 8-dimensional qudits. This uses a number of ideas from above, along with replacing the "query Hamiltonian" of [Ambainis, CCC 2014] with a new "sifter" construction.