Search papers, labs, and topics across Lattice.
This paper introduces Kolmogorov-Arnold Reservoir Computing (KARC), which enhances traditional reservoir computing by utilizing explicit basis-function expansions derived from the Kolmogorov-Arnold representation theorem. KARC maintains the expressive capacity of Kolmogorov-Arnold networks while enabling efficient closed-form training, addressing the limitations of conventional reservoir computing methods that struggle with long-range dependencies and hyperparameter sensitivity. The results demonstrate that KARC outperforms existing reservoir computing approaches on complex benchmarks, including partial differential equations, and can be integrated with generative diffusion models for improved text-to-image generation.
KARC not only preserves the expressive power of Kolmogorov-Arnold networks but also achieves superior performance on complex dynamical systems with efficient training.
Reservoir computing offers a lightweight framework for forecasting dynamical systems but may struggle to capture long-range dependencies due to limited representational capacity. Conventional reservoir computing recurrently uses trainable reservoirs with hyperparameter sensitivity, while the next-generation reservoir computing removes recurrence at the cost of rapidly growing feature dimensions. Here, we develop Kolmogorov-Arnold Reservoir Computing (KARC), which replaces reservoirs with explicit basis-function expansions inspired by the Kolmogorov-Arnold representation theorem. We rigorously show that KARC is a lightweight design of Kolmogorov-Arnold networks (KANs), preserving the potential expressive capacity of KANs while admitting efficient closed-form training of reservoir computing. At comparable cost, KARC outperforms existing reservoir computing methods on challenging benchmarks including partial differential equations. It can also be integrated with generative diffusion models for text-to-image generation. This work thus establishes a principled bridge between reservoir computing and KANs, enabling efficient and high-fidelity dynamical system forecasting.
