Metodebevisregistrering

Kolmogorov-Arnold Networks

Kolmogorov-Arnold Networks (KAN) is a neural network architecture introduced by Liu et al. in 2024 that replaces linear transformations with learned univariate functions on edges. Inspired by the Kolmogorov-Arnold representation theorem, KAN achieves superior function approximation with fewer parameters than traditional MLPs, offering potential efficiency gains and improved interpretability.

Sources recorded, not reviewed

Kilderegistrering

Citater kopieret ordret fra metodens kilderegistrering. Ingen påstandsniveauverifikation er udledt heraf.

KAN: Kolmogorov-Arnold Networks

Taksonomisk metoderegistrering · ml-model / deep-learning

Liu, Z., Wang, Y., Vaidya, S., Ruehle, F., Halverson, J., Soljačić, M., Hou, T. Y., & Tegmark, M. (2024). KAN: Kolmogorov-Arnold Networks. arXiv preprint arXiv:2404.19756. · URL

Åbn fuld metode

Kuraterede påstande

Påstande gemt i bevis-loggen, hver med sin egen vurdering.

Ingen kuraterede påstande endnu

Denne visning opfinder ikke en påstandsvurdering, når loggen ingen har.

Relaterede metoder

Genereret fra metodegrafen og vist som maskinelt foreslåede relationer — ingen bevispåstand er udledt.

Same method familyMamba (State Space Model)machine-suggested · Relational suggestion, not evidence.Same method familyMasked Autoencodersmachine-suggested · Relational suggestion, not evidence.Same method familyNeural Radiance Fields (NeRF)machine-suggested · Relational suggestion, not evidence.Same method familyVision Transformermachine-suggested · Relational suggestion, not evidence.