Gated QKAN-FWP: Scalable Quantum-inspired Sequence Learning

The article introduces the gated QKAN-FWP, a novel fast-weight framework that combines Fast Weight Programmers (FWPs) with Quantum-inspired Kolmogorov-Arnold Networks (QKAN) using single-qubit data re-uploading circuits. This 12.5k-parameter model outperforms classical recurrent baselines, including LSTMs and WaveNet-LSTMs, in long-horizon forecasting tasks while maintaining compatibility with noisy intermediate-scale quantum (NISQ) devices, achieving forecasting accuracy within 0.1% relative MSE of noiseless simulations. This work is significant for practitioners as it offers a scalable and efficient approach to sequence modeling that leverages quantum-inspired techniques.