Search papers, labs, and topics across Lattice.
This paper investigates the necessity of reconstruction in quantum machine learning (QML) by introducing a late fusion approach that combines outputs from independently trained quantum subcircuits, significantly reducing the classical sampling overhead. By employing a quantumness dial to balance between pure fusion and full reconstruction, the authors demonstrate that late fusion achieves comparable accuracy to full reconstruction while operating at exponentially lower costs and exhibiting greater robustness to noise. The findings suggest that late fusion can serve as an efficient alternative to traditional reconstruction methods in circuit-cutting QML applications.
Late fusion in quantum machine learning achieves near-identical accuracy to full reconstruction while slashing costs and enhancing noise resilience.
Circuit cutting lets a large quantum neural network (QNN) run as independent subcircuits on small devices, but rebuilding its outputs by reconstruction carries a classical sampling overhead exponential in the number of cuts - the dominant runtime cost in prior work. We ask whether, for machine-learning tasks, this step is necessary, and replace it with late fusion: each subcircuit is trained and measured independently, and a small classical head combines their outputs - a linear-cost, decision-level combination borrowed from multimodal learning. To characterize the trade-off we introduce a quantumness dial $Q$, a tunable reconstruction budget interpolating from pure fusion to full reconstruction, and a cut-entanglement diagnostic that indicates how much reconstruction a task needs (Spearman $\rho=0.59$ over $104$ runs). Across synthetic and standard datasets, independently trained late fusion matches full reconstruction accuracy within $0.04$ at every point of the controlled sweep and on every classical benchmark, at exponentially lower cost; it is also markedly more robust to shot and device noise. Controlled entangled-data experiments locate the boundary where fusion must fail. We do not claim advantage over classical machine learning - consistent with recent benchmarking, quantum offers no accuracy edge on these datasets. Late fusion is thus an efficient, noise-robust, self-characterizing alternative to reconstruction for circuit-cutting QML.
