Learning efficient decoders for quasi-chaotic quantum scramblers

Lorenzo Leone,Salvatore F.E. Oliviero,Seth Lloyd,Alioscia Hamma

DOI: https://doi.org/10.1103/PhysRevA.109.022429

2024-03-05

Abstract:Scrambling of quantum information is an important feature at the root of randomization and benchmarking protocols, the onset of quantum chaos, and black-hole physics. Unscrambling this information is possible given perfect knowledge of the scrambler [<a class="link-https" data-arxiv-id="1710.03363" href="https://arxiv.org/abs/1710.03363">arXiv:1710.03363</a>.]. We show that one can retrieve the scrambled information even without any previous knowledge of the scrambler, by a learning algorithm that allows the building of an efficient decoder. Remarkably, the decoder is classical in the sense that it can be efficiently represented on a classical computer as a Clifford operator. It is striking that a classical decoder can retrieve with fidelity one all the information scrambled by a random unitary that cannot be efficiently simulated on a classical computer, as long as there is no full-fledged quantum chaos. This result shows that one can learn the salient properties of quantum unitaries in a classical form, and sheds a new light on the meaning of quantum chaos. Furthermore, we obtain results concerning the algebraic structure of $t$-doped Clifford circuits, i.e., Clifford circuits containing t non-Clifford gates, their gate complexity, and learnability that are of independent interest. In particular, we show that a $t$-doped Clifford circuit $U_t$ can be decomposed into two Clifford circuits $U_{0},U^{\prime}_0$ that sandwich a local unitary operator $u_t$, i.e., $U_t=U_{0} u_{t}U_{0}^{\prime}$. The local unitary operator $u_t$ contains $t$ non-Clifford gates and acts nontrivially on at most $t$ qubits. As simple corollaries, the gate complexity of the $t$-doped Clifford circuit $U_t$ is $O(n^2+t^3)$, and it admits a efficient process tomography using $\mathrm{poly}(n,2^t)$ resources.

Quantum Physics