Approximate unitary $t$-designs by short random quantum circuits using nearest-neighbor and long-range gates

TL;DR

This paper proves that short local random quantum circuits on a D-dimensional lattice form approximate t-designs, improving bounds on depth and supporting quantum advantage in classical sampling hardness.

Contribution

It establishes that poly(t)·n^{1/D}-depth local random circuits are approximate t-designs, improving previous bounds and confirming conjectures about depth sufficiency for anti-concentration.

Findings

Poly(t)·n^{1/D} depth circuits form approximate t-designs.

Improved bounds for scrambling and decoupling in local circuits.

Depth O(√n) suffices for anti-concentration in 2D circuits.

Abstract

We prove that $poly(t) \cdot n^{1/D}$-depth local random quantum circuits with two qudit nearest-neighbor gates on a $D$-dimensional lattice with n qudits are approximate $t$-designs in various measures. These include the "monomial" measure, meaning that the monomials of a random circuit from this family have expectation close to the value that would result from the Haar measure. Previously, the best bound was $poly(t)\cdot n$ due to Brandao-Harrow-Horodecki (BHH) for $D=1$. We also improve the "scrambling" and "decoupling" bounds for spatially local random circuits due to Brown and Fawzi. One consequence of our result is that assuming the polynomial hierarchy (PH) is infinite and that certain counting problems are $\#P$-hard on average, sampling within total variation distance from these circuits is hard for classical computers. Previously, exact sampling from the outputs of even constant-depth quantum circuits was known to be hard for classical computers under the assumption that PH is infinite. However, to show the hardness of approximate sampling using this strategy requires that the quantum circuits have a property called "anti-concentration", meaning roughly that the output has near-maximal entropy. Unitary 2-designs have the desired anti-concentration property. Thus our result improves the required depth for this level of anti-concentration from linear depth to a sub-linear value, depending on the geometry of the interactions. This is relevant to a recent proposal by the Google Quantum AI group to perform such a sampling task with 49 qubits on a two-dimensional lattice and confirms their conjecture that $O(\sqrt n)$ depth suffices for anti-concentration. We also prove that anti-concentration is possible in depth O(log(n) loglog(n)) using a different model.