Benchmarking the computational power of quantum computers
本文提出了一种新的基准QUOPS，用于直接跨平台测量量子计算机执行最大相关电路的能力及其速度，以评估不同量子处理器的计算能力。
0 citationsRead paper
Selected work
Representative Papers
Towards unsupervised representation learning for quantum data: quantum models with inference and generation
研究提出了一种用于量子数据的无监督表示学习框架，通过推断和生成方法处理量子状态数据，克服了量子态缺乏经典概率论中标准分解属性的问题。
0 citationsRead paper
Automated near-term quantum algorithm discovery for molecular ground states
This work addresses the challenges of high resource consumption and algorithmic complexity in current quantum approaches to molecular ground-state estimation. We propose Hive, an AI-driven framework that integrates large language models with distributed evolutionary algorithms to enable, for the first time, the automated discovery of quantum heuristic algorithms. Applied to LiH, H₂O, and F₂ molecules, Hive generates highly efficient quantum circuits that substantially reduce quantum resource requirements. Through interpretability analysis, we uncover the mechanisms underlying performance gains. The discovered algorithms are experimentally validated on the Quantinuum H2 quantum processor, achieving chemical accuracy with fewer resources and precisely identifying the minimal hardware specifications necessary to reach this benchmark.
0 citationsRead paper
Toward Generative Quantum Utility via Correlation-Complexity Map
This study addresses the challenge of aligning real-world data distributions with the structural constraints of IQP-type quantum generative models. To this end, the authors propose a correlation-complexity mapping framework that introduces, for the first time, a two-dimensional criterion composed of the Quantum Correlation Likeness Index (QCLI) and the Classical Correlation Complexity Index (CCI) to assess data-model compatibility and guide model design. The methodology integrates reversible floating-point-to-bitstring encoding, low-dimensional latent trajectory interpolation, Walsh–Hadamard spectral analysis, Jensen–Shannon divergence, Chow–Liu tree approximation, and maximum mean discrepancy (MMD) optimization. Validation on turbulent flow data demonstrates that datasets exhibiting high QCLI and high CCI enable IQP models to achieve distributional fidelity comparable to RBMs and DCGANs, using fewer training samples and smaller latent architectures.
0 citationsRead paper
Causal and Compositional Abstraction
This work addresses the fundamental challenge of preserving causal structure during abstraction from low-level to high-level models—a key issue in scientific modeling, causal inference, and interpretable AI. Leveraging category theory, it formalizes causal abstraction as natural transformations, unifying existing frameworks and distinguishing upward from downward abstraction, with the latter—centered on high-level query mappings—shown to be more foundational. The paper introduces the novel notion of component-wise abstraction to strengthen mechanistic, constructive causal modeling and extends this framework for the first time to quantum combinatorial circuits, opening new avenues for interpretable quantum AI. By integrating intervention semantics and query mappings within compositional models in monoidal, cd-, or Markov categories, it establishes a unified theory of causal abstraction, proves a characterization theorem for mechanism-level abstraction, and demonstrates an initial abstraction mapping between classical causal models and quantum circuits.
0 citationsRead paper
Recent publications
Latest Papers
Benchmarking the computational power of quantum computers
本文提出了一种新的基准QUOPS，用于直接跨平台测量量子计算机执行最大相关电路的能力及其速度，以评估不同量子处理器的计算能力。
0 citationsRead paper
Towards unsupervised representation learning for quantum data: quantum models with inference and generation
研究提出了一种用于量子数据的无监督表示学习框架，通过推断和生成方法处理量子状态数据，克服了量子态缺乏经典概率论中标准分解属性的问题。
0 citationsRead paper
Automated near-term quantum algorithm discovery for molecular ground states
This work addresses the challenges of high resource consumption and algorithmic complexity in current quantum approaches to molecular ground-state estimation. We propose Hive, an AI-driven framework that integrates large language models with distributed evolutionary algorithms to enable, for the first time, the automated discovery of quantum heuristic algorithms. Applied to LiH, H₂O, and F₂ molecules, Hive generates highly efficient quantum circuits that substantially reduce quantum resource requirements. Through interpretability analysis, we uncover the mechanisms underlying performance gains. The discovered algorithms are experimentally validated on the Quantinuum H2 quantum processor, achieving chemical accuracy with fewer resources and precisely identifying the minimal hardware specifications necessary to reach this benchmark.
0 citationsRead paper
Toward Generative Quantum Utility via Correlation-Complexity Map
This study addresses the challenge of aligning real-world data distributions with the structural constraints of IQP-type quantum generative models. To this end, the authors propose a correlation-complexity mapping framework that introduces, for the first time, a two-dimensional criterion composed of the Quantum Correlation Likeness Index (QCLI) and the Classical Correlation Complexity Index (CCI) to assess data-model compatibility and guide model design. The methodology integrates reversible floating-point-to-bitstring encoding, low-dimensional latent trajectory interpolation, Walsh–Hadamard spectral analysis, Jensen–Shannon divergence, Chow–Liu tree approximation, and maximum mean discrepancy (MMD) optimization. Validation on turbulent flow data demonstrates that datasets exhibiting high QCLI and high CCI enable IQP models to achieve distributional fidelity comparable to RBMs and DCGANs, using fewer training samples and smaller latent architectures.
0 citationsRead paper
Causal and Compositional Abstraction
This work addresses the fundamental challenge of preserving causal structure during abstraction from low-level to high-level models—a key issue in scientific modeling, causal inference, and interpretable AI. Leveraging category theory, it formalizes causal abstraction as natural transformations, unifying existing frameworks and distinguishing upward from downward abstraction, with the latter—centered on high-level query mappings—shown to be more foundational. The paper introduces the novel notion of component-wise abstraction to strengthen mechanistic, constructive causal modeling and extends this framework for the first time to quantum combinatorial circuits, opening new avenues for interpretable quantum AI. By integrating intervention semantics and query mappings within compositional models in monoidal, cd-, or Markov categories, it establishes a unified theory of causal abstraction, proves a characterization theorem for mechanism-level abstraction, and demonstrates an initial abstraction mapping between classical causal models and quantum circuits.
0 citationsRead paper