Abstract:We prove that random quantum circuits on any geometry, including a 1D line, can form approximate unitary designs over $n$ qubits in $\log n$ depth. In a similar manner, we construct pseudorandom unitaries (PRUs) in 1D circuits in $\text{poly} \log n $ depth, and in all-to-all-connected circuits in $\text{poly} \log \log n $ depth. In all three cases, the $n$ dependence is optimal and improves exponentially over known results. These shallow quantum circuits have low complexity and create only short-range entanglement, yet are indistinguishable from unitaries with exponential complexity. Our construction glues local random unitaries on $\log n$-sized or $\text{poly} \log n$-sized patches of qubits to form a global random unitary on all $n$ qubits. In the case of designs, the local unitaries are drawn from existing constructions of approximate unitary $k$-designs, and hence also inherit an optimal scaling in $k$. In the case of PRUs, the local unitaries are drawn from existing unitary ensembles conjectured to form PRUs. Applications of our results include proving that classical shadows with 1D log-depth Clifford circuits are as powerful as those with deep circuits, demonstrating superpolynomial quantum advantage in learning low-complexity physical systems, and establishing quantum hardness for recognizing phases of matter with topological order.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to generate random unitaries with extremely low depth in quantum circuits. Specifically, the authors prove that on any geometric structure, including one - dimensional linear structures, random quantum circuits can form approximate unitary designs within logarithmic depth and construct pseudorandom unitaries (PRUs) within polylogarithmic depth. These results represent an exponential improvement in circuit depth compared to known results, and in all three cases, the dependence on the number of qubits $n$ is optimal. ### Main Contributions 1. **Approximate Unitary Designs**: - The authors prove that on any geometric structure, approximate $k$-designs can be constructed by combining local random unitary operations, with a circuit depth of $\tilde{\mathcal{O}}(k)\cdot\log(n / \epsilon)$. - For one - dimensional circuits, without the need for ancillary qubits, the circuit depth is $\mathcal{O}(\log(n / \epsilon)\cdot k \text{polylog}(k))$. - For fully - connected circuits, when using $\mathcal{O}(n\log(n / \epsilon))$ ancillary qubits and $k\leq3$, the circuit depth is $\mathcal{O}(\log\log(n / \epsilon))$. 2. **Pseudorandom Unitaries (PRUs)**: - The authors show that on any geometric structure, $n$-qubit pseudorandom unitary operations can be generated within polylogarithmic depth $\text{polylog}(n)$ by combining local pseudorandom unitary operations. - For fully - connected circuits, the circuit depth can be further reduced to $\text{polyloglog}(n)$. ### Applications 1. **Provably - Efficient Shallow Classical Shadows**: - By using shallow random Clifford circuits instead of traditional linear - depth Clifford circuits, efficient classical shadow estimation can be achieved while maintaining the same sample complexity. - This method can significantly reduce the noise impact in experiments, making it possible to achieve high - fidelity shadow estimation on more qubits. 2. **Quantum Hardness of Recognizing Topological Order**: - The authors prove that recognizing topological order is a quantum - computing - hard problem at any polylogarithmic depth. - This result is of great significance for topological - order detection in materials science and atomic, molecular, and optical experiments. 3. **Quantum Advantage for Learning Low - Complexity Quantum Systems**: - The authors show the advantage of quantum learning algorithms over classical algorithms in low - complexity systems. - In particular, in the task of distinguishing between random unitary processes and completely depolarizing channels, quantum algorithms can solve the problem efficiently, while classical algorithms require a super - polynomial number of queries. 4. **The Power of Time - Reversal in Learning**: - The authors use time - reversal techniques to show that modern quantum experiments are more efficient in learning quantum - dynamical properties than traditional experiments.

Random unitaries in extremely low depth

Randomized Benchmarking Using Unitary t-Design for Average Fidelity Estimation of Practical Quantum Circuit

Local random quantum circuits form approximate designs on arbitrary architectures

Approximate Unitary $k$-Designs from Shallow, Low-Communication Circuits

Efficient unitary designs and pseudorandom unitaries from permutations

How to Construct Random Unitaries

Unitary k-designs from random number-conserving quantum circuits

Simple constructions of linear-depth t-designs and pseudorandom unitaries

More global randomness from less random local gates

Efficient unitary designs with a system-size independent number of non-Clifford gates

Quantum advantage from measurement-induced entanglement in random shallow circuits

Asymptotically Optimal Circuit Depth for Quantum State Preparation and General Unitary Synthesis

Pseudorandom unitaries are neither real nor sparse nor noise-robust

Efficient Unitary T-designs from Random Sums

Pseudorandom unitaries with non-adaptive security

Anti-Concentration for the Unitary Haar Measure and Applications to Random Quantum Circuits

Measurement-induced entanglement and complexity in random constant-depth 2D quantum circuits

Shallow Shadows: Expectation Estimation Using Low-Depth Random Clifford Circuits

Fast pseudorandom quantum state generators via inflationary quantum gates

Query and Depth Upper Bounds for Quantum Unitaries via Grover Search

Unitary Designs from Random Symmetric Quantum Circuits