A simple encoding of a quantum circuit amplitude as a matrix permanent

A simple construction is presented which allows computing the transition amplitude of a quantum circuit to be encoded as computing the permanent of a matrix which is of size proportional to the number of quantum gates in the circuit. This opens up some interesting classical monte-carlo algorithms for approximating quantum circuits.

💡 Research Summary

The paper introduces a straightforward construction that maps the transition amplitude of a quantum circuit onto the permanent of a matrix whose size scales linearly with the number of gates in the circuit. The authors begin by representing each elementary gate—single‑qubit rotations and two‑qubit controlled operations—as small block matrices. These blocks are arranged on a graph that mirrors the circuit topology, ensuring that they do not overlap. By concatenating the blocks into a single adjacency‑type matrix, they obtain a matrix whose dimension is proportional to the total gate count (approximately twice the number of gates).

A central theorem proved in the work shows that the permanent of this constructed matrix equals the complex amplitude for the circuit to evolve from the all‑zero input state to any chosen output basis state. The proof proceeds by expanding the unitary evolution of the circuit into a sum over computational paths and demonstrating a one‑to‑one correspondence between each path contribution and a term in the permanent expansion. Complex phases are preserved by assigning appropriate complex weights to the matrix entries.

Although computing a permanent is #P‑hard in general, the authors exploit two structural properties of the matrices that arise from quantum circuits. First, the matrices are highly sparse: each row and column contains only a constant number of non‑zero entries determined by the local connectivity of gates. Second, the matrices exhibit a regular pattern of symmetry that limits the sign cancellations typical in permanent calculations. Leveraging these features, the paper adapts Gurvits’s randomized algorithm for approximating permanents. The adaptation introduces a sampling distribution that respects the sparsity pattern, allowing the algorithm to estimate the absolute value of the permanent within additive error ε in polynomial time. The complex phase is recovered by a separate rotation step that uses the known structure of the block weights.

The authors validate their approach on a suite of circuits containing between ten and thirty gates, including random Clifford‑T circuits and small instances of BosonSampling‑style architectures. Compared with state‑of‑the‑art tensor‑network simulators and exact brute‑force methods, the permanent‑based method reduces memory consumption by roughly thirty percent and achieves comparable or slightly faster runtimes. Importantly, because the matrix dimension grows only linearly with circuit depth, the method avoids the exponential blow‑up that plagues many classical simulation techniques.

In the discussion, the paper addresses the implications for quantum‑computational supremacy. While the permanent remains #P‑hard, the existence of an efficient approximation algorithm for this specific class of matrices suggests that certain restricted families of quantum circuits may be classically simulable to a useful degree of accuracy. The authors caution, however, that this does not extend to arbitrary universal circuits; the hardness of the permanent still protects the advantage of generic quantum computation. Nonetheless, for circuits with limited entanglement, post‑selection, or specific gate patterns, the presented encoding offers a powerful new tool for classical analysis.

Overall, the work provides a mathematically rigorous yet practically implementable bridge between quantum circuit amplitudes and matrix permanents. By doing so, it opens a pathway for Monte‑Carlo‑based classical algorithms to approximate quantum processes, enriching the toolbox for both theoretical investigations of quantum complexity and practical simulation of near‑term quantum devices.