AI 中文总结
研究为机器学习设计可扩展量子电路的障碍及解决方法，提出酉砖墙架构，通过调整粒子数k平衡经典模拟难度与训练成本，利用多层并行参数移位规则等提升效率，确立可训练性，酉蝴蝶变体也有类似成果。
AI 中文摘要
为机器学习设计可扩展的参数化量子电路面临三个基本障碍：阻碍基于梯度训练的贫瘠高原、缺乏对学习函数类经典困难性的可证保证以及每个梯度步骤中过高的电路评估成本。我们提出酉砖墙：一种用于最近邻硬件的k粒子费米子架构，将可重构分束器门与交错的单量子比特相位门和非高斯魔法态编码相结合，其中粒子数k是权衡经典模拟难度与训练成本的可调参数。该架构具有动态李代数\(\mathfrak{u}(n)\)，通过吉文斯旋转直接参数化\(U(n)\)，实现哈尔初始化。两体相关器读出实现梯度方差\(\Theta(k^2/n^5)\)。经典硬度由粒子数k控制。一种多层并行参数移位规则从每个梯度步骤的k(8n + 4)次电路评估中计算所有\(O(n^2)\)个梯度，比标准参数移位规则所需的\(3n^2\)次评估减少了\(3n/(8k)\)倍。酉蝴蝶变体针对全对全硬件，深度为\(2\log n\)，有\(n\log n\)个参数。它在每个梯度步骤8klogn次评估时实现了类似的硬度保证，同样减少了\(3n/(8k)\)倍。其可训练性在两个层面得到确立：不存在指数级贫瘠高原是无条件的，而尖锐的\(\Theta(k^2/n^5)\)速率在两粒子近似2设计猜想下成立。
英文摘要
Designing scalable parameterized quantum circuits for machine learning faces three obstacles: barren plateaus, the absence of guarantees that the learned function class is classically hard, and prohibitive circuit evaluations per gradient step. We propose the unitary brick-wall: a $k$-particle fermionic architecture for nearest-neighbor hardware, combining Reconfigurable Beam Splitter gates with interleaved single-qubit phase gates and a non-Gaussian magic-state encoding, where $k$ is a tunable dial trading classical simulation hardness against training cost. Trainable. The brick-wall has dynamical Lie algebra $\mathfrak{u}(n)$ and is surjective onto $U(n)$ via Givens rotations, enabling Haar initialization. Two-body correlator readouts achieve gradient variance $Θ(k^3/n^5)$, polynomial in $n$ throughout $n-2k=Ω(n)$. Expressive. Classical hardness is controlled by $k$: best-known classical sampling algorithms run in time $2^{Θ(k)}\mathrm{poly}(n)$, worst-case #P-hardness holds from $k=n^ε$, and the average-case machinery of Fermion Sampling applies at $k=Θ(n)$. At our operating point $k=60$, best-known classical simulation exceeds $10^{24}$ operations at every $n$. Efficient. A multi-layer parallel parameter-shift rule computes all $O(n^2)$ gradients from $4kn$ circuit evaluations per gradient step, a factor $n/k$ reduction over the $4n^2$ evaluations of the standard rule, growing linearly with $n$ at fixed $k$. The unitary butterfly variant targets all-to-all hardware, with depth $2\log n$ and $(3/2)n\log n$ parameters, similar hardness guarantees, and $4k\log n$ evaluations per gradient step -- the same factor-$n/k$ reduction. Its trainability holds at two levels: absence of exponential barren plateaus is unconditional, while the sharp $Θ(k^3/n^5)$ rate holds under a two-particle approximate-2-design conjecture.
Comments45 pages. v2: Exact closed-form gradient variance with improved $Θ(k^3/n^5)$ rate; improved parameter-shift count. The simulability proposition now covers arbitrary fixed body number via $r$-RDM propagation (we thank Erfan Amidi), making triplet-block two-body readouts polynomial-time classically; that encoding is removed. Sampling hardness unchanged. Extended ML pipeline presentation
