CVKAN: Complex-Valued Kolmogorov-Arnold Networks
Matthias Wolff, Florian Eilers, Xiaoyi Jiang · 2025
Read CVKAN: Complex-Valued Kolmogorov-Arnold Networks by Matthias Wolff, Florian Eilers, Xiaoyi Jiang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-02417.txt.
Description
Matthias Wolff, Florian Eilers, and Xiaoyi Jiang build complex-valued Kolmogorov-Arnold networks in a paper from 2025. KANs are neural networks prized for being readable: their parts map to plain formulas. Classic neural nets for complex numbers handle signals like audio and radar well, but lack that clarity.
CVKAN merges both strengths by moving the whole KAN design into the complex domain. Tests fit symbolic complex functions and formulas from physics. A harder test came from knot theory data. The complex network proved more stable than real-valued KANs. It matched or beat them with fewer parameters and a shallower shape. Readable models for complex signals serve engineers who must explain their tools.
Catalog record
| CREATOR | Matthias Wolff, Florian Eilers, Xiaoyi Jiang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-02417