Bottom-Up Design of Quantum Optical Experiments Using Discrete Generative Models
Abstract
Designing quantum optical experiments requires searching over discrete circuit topologies and continuous parameters, often with multiple realizations of the same target state. Graph-based methods commonly address this problem by optimizing a dense graph and pruning it toward a single circuit. We introduce \texttt{Grinch}, a bottom-up reward-driven generative framework that learns to sample optical graphs directly from fidelity-based rewards without relying on a pre-existing training dataset. We demonstrate the framework on multipartite entangled states, graph and cluster states, and nonlocal photonic-gate targets, obtaining multiple high-fidelity optical graphs. We identify asymptotic solutions for states that cannot be generated exactly by graphs, as well as nonlocal Toffoli gate constructions requiring two ancillas. Our work presents hardware-constraint objectives, which lead to alternative constructions of CNOT and Toffoli gates when connections between ancilla nodes are restricted. These findings illustrate a reward-driven approach to quantum optical inverse design, exploring alternative circuit topologies while directly incorporating connectivity constraints into the search process.
Related
- Flexible Inverse Design of Optical Multilayer Coatings Using Discrete-Continuous Flow Matching
- Enhancing Quantum Metrology with High-order Fisher Information and Experiments
- The convergent laboratory: when AI reasoning, autonomous experiments, high performance and quantum computing reshape chemistry
- Simulating Quantum Physics Experiments One Event at a Time
- Quantum number conservation: a tool in the design and analysis of high energy experiments
BibTeX
@misc{huidobromeezs2026bottomup,
title = {{Bottom-Up Design of Quantum Optical Experiments Using Discrete Generative Models}},
author = {Isaac L. Huidobro-Meezs and Simón Paiva-Ortega and Rodrigo A. Vargas-Hernández},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.08073},
eprint = {2609.08073},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.08073},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Bottom-Up Design of Quantum Optical Experiments Using Discrete Generative Models
AU - Isaac L. Huidobro-Meezs
AU - Simón Paiva-Ortega
AU - Rodrigo A. Vargas-Hernández
PY - 2026
DA - 2026/09/08
PB - arXiv
DO - 10.48550/arxiv.2609.08073
UR - https://arxiv.org/abs/2609.08073
AB - Designing quantum optical experiments requires searching over discrete circuit topologies and continuous parameters, often with multiple realizations of the same target state. Graph-based methods commonly address this problem by optimizing a dense graph and pruning it toward a single circuit. We introduce \texttt{Grinch}, a bottom-up reward-driven generative framework that learns to sample optical graphs directly from fidelity-based rewards without relying on a pre-existing training dataset. We demonstrate the framework on multipartite entangled states, graph and cluster states, and nonlocal photonic-gate targets, obtaining multiple high-fidelity optical graphs. We identify asymptotic solutions for states that cannot be generated exactly by graphs, as well as nonlocal Toffoli gate constructions requiring two ancillas. Our work presents hardware-constraint objectives, which lead to alternative constructions of CNOT and Toffoli gates when connections between ancilla nodes are restricted. These findings illustrate a reward-driven approach to quantum optical inverse design, exploring alternative circuit topologies while directly incorporating connectivity constraints into the search process.
ER -