analiza aiqbench
This paper establishes rigorous mathematical bounds for Gaussian boson sampling by analyzing the uniform hiding properties of Haar block transpose Gram matrices, directly advancing our understanding of quantum computational hardness. The explicit error analysis provides a critical random matrix replacement component that strengthens theoretical arguments regarding the complexity of specific quantum photonic architectures.
Uniform Hiding of Haar Block Transpose Gram Matrices
arXiv:2609.01008v1 Announce Type: new Abstract: Gaussian boson sampling with equally squeezed active inputs assigns collision free probabilities through a complex symmetric transpose Gram matrix formed from a rectangular block of a Haar interferometer. We prove a finite total variation comparison with the corresponding complex Gaussian transpose Gram law. The error bound is explicit, quadratic in the number of selected output modes, inversely proportional to the interferometer size, and uniform in the number of squeezed active inputs. The proof combines a centered circular orthogonal ensemble score analysis in the dense regime with a rectangular relative entropy bound. This result supplies a random matrix replacement component of Gaussian boson sampling hardness arguments.
Abstract:Gaussian boson sampling with equally squeezed active inputs assigns collision free probabilities through a complex symmetric transpose Gram matrix formed from a rectangular block of a Haar interferometer. We prove a finite total variation comparison with the corresponding complex Gaussian transpose Gram law. The error bound is explicit, quadratic in the number of selected output modes, inversely proportional to the interferometer size, and uniform in the number of squeezed active inputs. The proof combines a centered circular orthogonal ensemble score analysis in the dense regime with a rectangular relative entropy bound. This result supplies a random matrix replacement component of Gaussian boson sampling hardness arguments.
arXiv:2609.01008 [quant-ph]
(or arXiv:2609.01008v1 [quant-ph] for this version)
https://doi.org/10.48550/arXiv.2609.01008 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 196 cuvinte
Subiecte conexe
Local Anticoncentration for Gaussian Boson Sampling via Conditional Wishart Geometry
arXiv · 2 oct. 2026
Productionizing TurboQuant on AMD GPUs for KV-Cache-Bound LLM Inference
AMD · 11 iun. 2026

Challenge: Hand coding weights for efficient sequence memorisation
AI Alignment Forum · 23 iul. 2026
Enabling DeepSeek-V4-Flash Training on AMD Instinct MI355X GPUs with Primus
AMD · 3 sept. 2026