Track the frontier, from condensed matter to cosmology.
Search papers, labs, and topics across Lattice.
To bypass the exponentially intractable likelihoods of quantum parameter estimation, this work pairs polynomial-cost classical simulators—such as tensor networks and Pauli propagation—with normalizing flows for amortized neural posterior estimation. Pre-training neural density estimators on simulated outputs amortizes the computational burden into a single up-front cost, allowing experimental measurement records to be mapped to full parameter posteriors in a single forward pass. Validated across Hamiltonian learning, quantum state tomography, and Pauli noise characterization, the method successfully scales to 81-qubit circuits and 735 parameters while leveraging posterior variance to diagnose parameter non-identifiability.
Quantum system characterization no longer requires intractable per-experiment likelihood evaluations: pairing efficient classical simulators with normalizing flows yields full parameter posteriors in a single forward pass on up to 81 qubits.
Models of quantum systems faithfully map system parameters to observations, but the inverse problem of parameter inference from measurement data presents a fundamental challenge: computationally intractable likelihoods due to an exponentially large Hilbert space. Here, we introduce simulation-based quantum system inference, a unified, likelihood-free framework that learns parameter posteriors directly from classical simulation data. The central idea is to pair polynomial-cost classical simulators, such as Pauli propagation and tensor networks, with normalizing flows or other neural density estimators for accurate, reusable inference. A single model, trained once, maps any new measurement record to its posterior in one forward pass---turning per-experiment inference into a fixed, up-front cost. We numerically demonstrate the framework's versatility across Pauli noise learning, quantum error mitigation, quantum state tomography, and Hamiltonian learning, with examples involving 81-qubit shallow circuits and 735-parameter inference. In each case, the approach yields accurate estimates of identifiable parameters, while posterior uncertainty provides additional diagnostics of non-identifiability and indicates where further characterization is needed. Our framework reduces data-acquisition requirements in quantum experiments and accelerates parameter inference, providing a practical route to characterizing and improving large-scale quantum systems.