Pulse-based optimization of quantum many-body states with Rydberg atoms in optical tweezer arrays

We explore a pulse-based variational quantum eigensolver (VQE) algorithm for Rydberg atoms in optical tweezer arrays and evaluate its performance on prototypical quantum spin models. We numerically demonstrate that the ground states of the one-dimensional antiferromagnetic Heisenberg model and the mixed-field Ising model can be accurately prepared using an adaptive update algorithm that randomly segments pulse sequences, for systems of up to ten qubits. Furthermore, we propose and validate a hybrid scheme that integrates this pulse-level analog quantum algorithm with a variational quantum gate approach, where digital quantum gates are approximated by optimized analog pulses. This enables efficient measurement of the cost function for target many-body Hamiltonians.