{"id":"arxiv/2608.16939","url":"https://arxiv.org/abs/2608.16939","tags":["quant-ph","cs.LG"],"score":0,"title":"SPSA Hyperparameter Tuning for Variational Quantum Natural Language Inference","domain":"quant-ph","source":"arxiv","authors":["Nayan D'Souza","Christopher J. Agostino"],"reviews":0,"subject":"quant-ph","summary":"Training variational quantum models requires choosing between parameter-shift gradients, which are exact but cost $O(P)$ forward evaluations, and simultaneous perturbation stochastic approximation (SPSA), which uses only two samples but produces high-variance estimates that can degrade optimisation on small supervised tasks. Whether the cheap gradient is usable depends on the variance that results from different choices of the SPSA perturbation scale, learning rate, and gain-decay schedule. We varied those quantities across a broad grid on a 6-qubit, 60-parameter QNLI classifier and compared…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-08-11","reactions":{}}