Adaptive Circuits for Quantum Chemistry

Adaptive Circuits for Quantum Chemistry

This PennyLane demo implements adaptive circuit construction for quantum chemistry, a family of methods (such as ADAPT-VQE) that build a problem-tailored ansatz on the fly instead of using a fixed, generic circuit, often yielding much shallower, more accurate chemistry calculations. Fixed ansaetze like full unitary coupled cluster can be unnecessarily deep, including many operators that contribute little; adaptive methods instead grow the ansatz one operator at a time, at each step selecting from a pool of candidate excitation gates the one whose gradient indicates it will lower the energy the most, adding it, re-optimizing, and repeating until convergence. The tutorial shows how to implement this in PennyLane: defining the molecular Hamiltonian and a pool of single- and double-excitation operators, computing the energy gradient with respect to each candidate to rank them, iteratively appending the most useful gates, and re-optimizing the parameters after each addition. It applies the method to a small molecule and shows that the adaptively-grown circuit reaches accurate energies with far fewer gates than a fixed ansatz. By building the circuit to fit the problem, the demo illustrates a powerful, resource-efficient approach to variational quantum chemistry in PennyLane.

Run it

pip install -r requirements.txt
python demo.py

Source and license

Imported from demonstrations_v2/tutorial_adaptive_circuits/demo.py in PennyLaneAI/demos at c52c0abeb5122218aa96b38eea848864cce7323f, under the Apache License 2.0. Original authors: Xanadu and the PennyLane community. The upstream LICENSE is included alongside this example.