Back to AI News
arxiv

Hardware-Trained Generative Quantum Eigensolver Scales to 70 Qubits for Organotin Hydroxides

Researchers trained generative quantum eigensolver policies, transformers conditioned on molecular descriptors, to produce state-preparation circuits for sample-based quantum diagonalization. Training moved from a noiseless simulator to the IBM Heron processor ibm_kobe in the loop at 46 and 70 qubits across four main runs using 1.2×10^8 shots. Policies trained on transition-metal complexes transferred without retraining to organotin hydroxide models from 32 to 70 qubits. After one hardware sampling round and classical ranking, 27 of 28 runs at 46–70 qubits were within 15 mHa of DMRG references, and selected circuits at 56–70 qubits used 34–72 times fewer two-qubit gates than transpiled LUCJ circuits.

AI Brief

It shows a concrete hardware-in-the-loop method for training generative circuit policies at 70 qubits, with transfer to new molecular systems and substantial circuit-efficiency gains.

arXiv:2611.13755v1 Announce Type: cross Abstract: Quantum hardware can provide feedback for training circuit generators beyond the reach of exact classical simulation. We train generative quantum eigensolver (GQE) policies, transformers conditioned on molecular descriptors, to generate the state-preparation circuits for sample-based quantum diagonalization (SQD). Training runs first on a noiseless simulator and then with the IBM Heron processor ibm_kobe inside the loop, at 46 and 70 qubits ($1.2 \times 10^8$ shots for the four main training runs). A classical CCSD(T) reference energy of the training molecules enters only an admission gate and two auxiliary reward terms. During hardware training, a reward penalty prevents collapse to circuits that sample only the Hartree-Fock determinant. Policies trained on transition-metal complexes are applied without retraining to Sn(OH)$_3$-Me and its hydrates, minimal models of organotin photoresists for extreme-ultraviolet lithography, from 32 to 70 qubits. After one round of hardware sampling, followed by classical ranking and recovery of the cached counts with the molecular Hamiltonian and no further QPU calls, 27 of 28 runs at 46-70 qubits are within 15 mHa of the density-matrix renormalization group (DMRG) reference (median 4.5 mHa), and in all eight runs with an untrained control the trained policy is more accurate. At 56-70 qubits the selected circuits use 34-72 times fewer two-qubit gates than the local unitary cluster Jastrow (LUCJ) circuit transpiled for the same device. These results demonstrate that circuit generators can be trained with quantum hardware at 70 qubits.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!