Gated QKAN-FWP: Scalable Quantum-inspired Sequence Learning

2026-05-07

AI Summary

This work addresses the poor scalability and high classical simulation cost of existing quantum fast weight programmers on noisy intermediate-scale quantum (NISQ) devices, which stem from their reliance on multi-qubit architectures. The authors propose a novel hybrid architecture that employs single-qubit data re-uploading circuits as learnable nonlinear activations (termed the DARUAN mechanism), integrated with a quantum-inspired Kolmogorov–Arnold network and a scalar-gated fast weight update rule. This design achieves adaptive memory kernels, geometric boundedness, and parallel gradient pathways. The approach substantially enhances NISQ compatibility and scalability, outperforming various classical recurrent models—despite using only 12.5k parameters compared to up to 13× more in baselines—on solar activity cycle forecasting (528-month input, 132-month prediction). It also attains near-ideal simulator accuracy on IonQ and IBM quantum processors, with relative mean squared error below 0.1%.

Abstract

Fast Weight Programmers (FWPs) encode temporal dependencies through dynamically updated parameters rather than recurrent hidden states. Quantum FWPs (QFWPs) extend this idea with variational quantum circuits (VQCs), but existing implementations rely on multi-qubit architectures that are difficult to scale on noisy intermediate-scale quantum (NISQ) devices and expensive to simulate classically. We propose gated QKAN-FWP, a fast-weight framework that integrates FWP with Quantum-inspired Kolmogorov-Arnold Network (QKAN) using single-qubit data re-uploading circuits as learnable nonlinear activation, known as DatA Re-Uploading ActivatioN (DARUAN). We further introduce a scalar-gated fast-weight update rule that stabilizes parameter evolution, supported by a theoretical analysis of its adaptive memory kernel, geometric boundedness, and parallelizable gradient paths. We evaluate the framework across time-series benchmarks, MiniGrid reinforcement learning, and highlight real-world solar cycle forecasting as our main practical result. In the long-horizon setting with 528-month input window and 132-month forecast horizon, our 12.5k-parameter model achieves lower scaled Mean Square Error (MSE), peak amplitude error, and peak timing error than a suite of classical recurrent baselines with up to 13x more parameters, including Long Short-Term Memory (LSTM) networks (25.9k-89.1k parameters), WaveNet-LSTM (167k), Vanilla recurrent neural network (11.5k), and a Modified Echo State Network (132k). To validate NISQ compatibility, we further deploy the trained fast programmer on IonQ and IBM Quantum processors, recovering forecasting accuracy within 0.1% relative MSE of the noiseless simulator at 1024 shots. These results position gated QKAN-FWP as a scalable, parameter-efficient, and NISQ-compatible approach to quantum-inspired sequence modeling.

Problem

• Quantum Fast Weight Programmers
• NISQ devices
• scalability
• sequence modeling
• parameter efficiency

Innovation

• Quantum-inspired learning
• Fast Weight Programmer
• Single-qubit re-uploading
• Scalable sequence modeling
• NISQ compatibility