TruKAN: Towards More Efficient Kolmogorov-Arnold Networks Using Truncated Power Functions
Ali Bayeh, Samira Sadaoui, Malek Mouhoub · 2026
Read TruKAN: Towards More Efficient Kolmogorov-Arnold Networks Using Truncated Power Functions by Ali Bayeh, Samira Sadaoui, Malek Mouhoub (2026) 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-2602-03879.txt.
Description
Ali Bayeh and 2 co-authors rework a young neural network design in their 2026 paper. Kolmogorov-Arnold Networks replace fixed activations with learnable curves, usually built from B-splines, which are smooth but costly. Their version, TruKAN, swaps in truncated power functions, a simpler family from spline theory.
Each layer pairs a truncated power term with a polynomial term, using shared or separate knots. Built into an EfficientNet-V2 vision framework, TruKAN beats rival KAN variants on accuracy, training time, and memory. The simpler curves also make the model easier to interpret than its spline-based cousins.
Catalog record
| CREATOR | Ali Bayeh, Samira Sadaoui, Malek Mouhoub |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 66 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-03879