Metraipolis › archive › dc_97499ef2b357

InKAN: B-Spline KANs via Truncated Power Form — arXiv:2609.01956

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.01956
License (as archived)
CC0-1.0
Archived
2026-10-17T04:59:21.180Z
SHA-256 of the archived text
b0e1ecf25ba653b32738a5b13a8f8ede7bc3110b311f55a8fe557664b965f4e0
Stable URL of this page
https://metraipolis.com/documento/dc_97499ef2b357

How to cite

InKAN: B-Spline KANs via Truncated Power Form — arXiv:2609.01956 — arXiv (en), https://arxiv.org/abs/2609.01956 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_97499ef2b357 (Metraipolis archive dc_97499ef2b357, sha256 b0e1ecf25ba653b32738a5b13a8f8ede7bc3110b311f55a8fe557664b965f4e0)
BibTeX
@misc{dc97499ef2b357,
  title = {InKAN: B-Spline KANs via Truncated Power Form — arXiv:2609.01956},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.01956},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_97499ef2b357; sha256 b0e1ecf25ba653b32738a5b13a8f8ede7bc3110b311f55a8fe557664b965f4e0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.01956. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Kolmogorov-Arnold Networks (KANs) place learnable B-spline activations on network edges rather than fixed activations on nodes. The standard Cox-de Boor recursion evaluates these activations through $k$ sequential passes for degree-$k$ splines, consuming over 90% of forward-pass time. InKAN replaces this recursion with the truncated power form, a classical result from approximation theory that expresses each uniform cubic B-spline as five $(x)_+^3$ terms at shifted knot positions. The resulting expression computes exact B-spline basis values: the same mathematical function as the Cox-de Boor recursion, evaluated without sequential passes. This paper documents three contributions: (1) an implementation structured for torch . compile fusion, eliminating recursion, span lookup, and scatter-gather operations; (2) a bounded-coordinate evaluation that clamps the normalized input to $[0, k{+}1]$, preventing the growth of cancellation error at large off-support coordinates; and (3) an open-source package (pip install inkan). In the tested configurations, InKAN has 2.8--3.5$\times$ lower forward-pass latency than the Cox-de Boor recursion. Partition-of-unity errors remain below $10^{-5}$ for grid sizes up to 200.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-03 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Numerical Analysis (math.NA).
Pagina del lavoro: https://arxiv.org/abs/2609.01956 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T04:09:43.297Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).