遇见数据集

QIRT Meta-Floor Google sim

收藏
Zenodo2025-09-06 更新2026-05-26 收录
官方服务:

资源简介:

This dataset and report document the execution and results of QIRT8v2, a symbolic clause alignment quantum circuit designed under the Quantum Information Refinement Theory (QIRT) framework. The benchmark was conducted on a Google Sycamore-style noisy simulator via Colab, emulating gate and readout noise similar to Google's quantum hardware. The test implements a 3-Exact Cover problem with 9 target qubits, 3 clause structures, and symbolic clause collapse logic without any cost Hamiltonian or penalty encoding. The simulation was executed over 1024 shots, with post-processing validating clause satisfaction and computing entropy of the output state distribution. Key findings include: Clause Satisfaction Rate (CSR): 5.08% Entropy: 4.21 bits Top valid symbolic solutions demonstrating coherent clause collapse despite injected noise This result confirms that QIRT’s symbolic scaffold is resilient even under Sycamore-type error models, demonstrating backend-agnostic collapse behavior and outperforming baseline expectations without Grover search or QEC overhead. Contents: qirt8v2_sycamore.qasm – Quantum circuit source notebook Author:Kristopher K. Knudsen (Axion Metrics LLC)Protected under U.S. Provisional Patent Application No. 63/826,473DOI (root benchmark): https://doi.org/10.5281/zenodo.15867661

本数据集与报告记录了QIRT8v2的运行情况与实验结果,QIRT8v2是基于量子信息精化理论(Quantum Information Refinement Theory, QIRT)框架设计的符号化子句对齐量子电路。本次基准测试通过谷歌Colab平台,在谷歌悬铃木(Sycamore)风格噪声模拟器上开展,模拟了与谷歌量子硬件相近的门操作噪声与读出噪声。 本次测试针对含9个目标量子比特、3个子句结构的3精确覆盖(3-Exact Cover)问题,采用符号化子句坍缩逻辑,未使用任何代价哈密顿量或惩罚编码。本次模拟共执行1024次采样(shots),后续处理环节对解的子句满足性进行验证,并计算输出态分布的熵值。 核心研究发现包括: 子句满足率(Clause Satisfaction Rate, CSR):5.08% 熵值:4.21比特 顶级有效符号化解表明,即便存在注入噪声,仍可实现连贯的子句坍缩。 该结果证实,即便在悬铃木风格的噪声模型下,QIRT的符号化架构仍具备鲁棒性,展现出与后端无关的坍缩特性,且无需格罗弗(Grover)搜索或量子纠错(Quantum Error Correction, QEC)开销即可超越基准预期。 文件内容: qirt8v2_sycamore.qasm —— 量子电路源代码笔记本 作者:克里斯托弗·K·克努森(Kristopher K. Knudsen)(Axion Metrics有限责任公司) 本作品受美国临时专利申请第63/826,473号保护 基准测试数字对象标识符(Digital Object Identifier, DOI):https://doi.org/10.5281/zenodo.15867661

提供机构:
Zenodo
创建时间:
2025-07-13
二维码
社区交流群
二维码
科研交流群
商业服务