A graph neural network trained on H4 and H6 predicts optimized orbitals for larger unseen H8-H12 systems with O(10-100) milli-Hartree energy errors and provides effective warm-starts for VQE optimization.
Optimized low-depth quantum circuits for molecular electronic structure using a separable-pair approx- imation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
quant-ph 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Transferable Machine Learning Approach to Predict Optimized Orbitals for Electronic Structure Problems
A graph neural network trained on H4 and H6 predicts optimized orbitals for larger unseen H8-H12 systems with O(10-100) milli-Hartree energy errors and provides effective warm-starts for VQE optimization.