Speakers
Description
Thermal equilibrium states of many-body Hamiltonians are central to studying quantum chaos, phases of matter, and quantum machine learning. Yet generating large ensembles of thermal states across different Hamiltonians remains prohibitively costly with existing methods. Here, we introduce the quantum thermal-drift channel, a new primitive that enables a measurement-controlled sampling algorithm to autonomously generate thermal states together with their Hamiltonian labels for general physical models. We prove that the algorithm is efficient: its total gate count scales polynomially with system size and quadratically with inverse temperature, establishing a polynomial resource bound for random thermal-state generation. For a two-dimensional transverse-field Ising model, we find that the sampled thermal states exhibit a crossover to Wigner--Dyson level statistics, demonstrating nontrivial chaotic correlations. As an application, we show that a variational quantum classifier trained on the generated dataset achieves near-optimal accuracy in predicting Hamiltonian properties of unseen states. These results establish a scalable, quantum-native route to thermodynamic simulation and quantum generative learning in many-body systems.
| Topic | Quantum algorithms for machine learning applications |
|---|---|
| Additional Topic | Hybrid quantum-classical approaches for learning and optimization |