发表机构
Reichman University; Dartmouth College; The Hebrew University of Jerusalem（赖赫曼大学; 达特茅斯学院; 耶路撒冷希伯来大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究通过受控消融实验，探究将PQCs作为残差边载模块集成到ResNet-50中用于神经解码的效果，发现骨干梯度训练投影的量子边载模型可提升准确率，测量引导训练能改善表征几何，且未实现量子计算优势。
AI 中文摘要
本研究将参数化量子电路（PQCs）作为残差边载模块集成到ResNet-50骨干网络中，用于31类神经群体解码任务——即从多神经元尖峰光栅中进行想象手写分类。在严格受控条件下（固定数据划分、随机种子和优化器），我们对比了四种模型变体：基线模型、输入投影冻结的量子边载模型、骨干梯度训练投影的量子边载模型，以及将角度编码与电路测量结果对齐的测量引导变体。骨干梯度变体在4个随机种子中的3个上提升了准确率（均值+0.19%，95%置信区间[-1.10%, +1.48%]），且在全部4个种子中均降低了与基线特征的线性中心核对齐（Linear CKA）相似度（差值=-0.025），表明表征发生了真实的结构重组。针对9种变体的消融研究确定，简单浅层架构是最有效且可复现的配置。测量引导训练可在不降低准确率的前提下持续改善表征几何结构。所有结果均基于4量子比特的无噪声态矢量模拟，该方案的选择是为了反映当前近期超导硬件的实际约束；本研究未声称取得超越经典方法的量子计算优势。
英文摘要
We study parameterized quantum circuits (PQCs) integrated as residual sidecar modules within a ResNet-50 backbone for 31-class neural population decoding---imagined handwriting classification from multi-neuron spike rasters. Under strictly controlled conditions (fixed data splits, seeds, and optimizer), we compare four model variants: baseline, quantum sidecar with frozen input projection, quantum sidecar with backbone-gradient-trained projection, and a measurement-guided variant that aligns angle encodings with circuit measurement outcomes. The backbone-gradient variant improves accuracy in 3/4 seeds (+0.19% mean, 95% CI [-1.10%, +1.48%]) and consistently reduces Linear CKA similarity to baseline features ($Δ=-0.025$, 4/4 seeds), indicating genuine structural reorganization of representations. A nine-variant ablation identifies simple shallow architectures as the most effective and reproducible configuration. Measurement-guided training consistently improves representation geometry without reducing accuracy. All results use noiseless statevector simulation on 4 qubits, a regime chosen to reflect the practical constraints of current near-term superconducting hardware; no quantum computational advantage over classical methods is claimed.
CommentsAccepted at the 5th International Workshop on Human Brain and Artificial Intelligence (HBAI 2026), IJCAI-ECAI 2026
