Pith. sign in

REVIEW 12 cited by

KAN or MLP: A Fairer Comparison

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2407.16674 v2 pith:J7HBEVK7 submitted 2024-07-23 cs.LG cs.AI

KAN or MLP: A Fairer Comparison

classification cs.LG cs.AI
keywords formularepresentationsymbolicb-splineperformancetaskscomparisonfairer
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

This paper does not introduce a novel method. Instead, it offers a fairer and more comprehensive comparison of KAN and MLP models across various tasks, including machine learning, computer vision, audio processing, natural language processing, and symbolic formula representation. Specifically, we control the number of parameters and FLOPs to compare the performance of KAN and MLP. Our main observation is that, except for symbolic formula representation tasks, MLP generally outperforms KAN. We also conduct ablation studies on KAN and find that its advantage in symbolic formula representation mainly stems from its B-spline activation function. When B-spline is applied to MLP, performance in symbolic formula representation significantly improves, surpassing or matching that of KAN. However, in other tasks where MLP already excels over KAN, B-spline does not substantially enhance MLP's performance. Furthermore, we find that KAN's forgetting issue is more severe than that of MLP in a standard class-incremental continual learning setting, which differs from the findings reported in the KAN paper. We hope these results provide insights for future research on KAN and other MLP alternatives. Project link: https://github.com/yu-rp/KANbeFair

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 12 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Low Complexity Kolmogorov-Arnold Network-based DPD for Analog RoF Fronthaul

    eess.SP 2026-06 unverdicted novelty 7.0

    An envelope time-delay Kolmogorov-Arnold Network (ETDKAN) for DPD in A-RoF systems matches MLP performance on ACLR/EVM with complexity near memory polynomials and yields 4-5 dB ACLR improvement in experiment.

  2. Neural Discovery of Memory and Nonlocal Kernels in Integro-Differential Equations with Constrained Kolmogorov--Arnold Networks

    cs.LG 2026-07 conditional novelty 6.0

    Hard-constrained Bernstein MC-KANs recover positive, monotone, convex memory/nonlocal kernels from sparse noisy IDE data more robustly than soft-penalized Cheb-KANs, especially in 2D.

  3. Adaptive RBF-KAN: A Comparative Evaluation of Dynamic Shape Parameters in Kolmogorov-Arnold Networks

    stat.ML 2026-05 unverdicted novelty 6.0

    Adaptive RBF-KAN adds multiple radial basis kernels and LOOCV-based shape initialization to FastKAN, with benchmark tests on 2D functions showing kernel-specific advantages for smooth, discontinuous, and oscillatory cases.

  4. Hardware-Oriented Inference Complexity of Kolmogorov-Arnold Networks

    cs.LG 2026-04 unverdicted novelty 6.0

    Derives generalized formulas for KAN inference complexity using RM, BOP, and NABS metrics across B-spline, GRBF, Chebyshev, and Fourier variants.

  5. Deep Learning Alternatives of the Kolmogorov Superposition Theorem

    cs.LG 2024-10 unverdicted novelty 6.0

    ActNet is a new KST-based neural network that outperforms KANs and competes with MLPs in PINN benchmarks for PDE simulation tasks.

  6. KAN-LSTM-Transformer Neural Networks, MFV and Cosmological Parameters

    astro-ph.CO 2026-07 conditional novelty 5.0

    KLT-Net reconstructs the SN Ia distance modulus non-parametrically; with MFV M_B and flat-ΛCDM Bayesian/Hessian inference it yields H0 ≈ 69.6 km s⁻¹ Mpc⁻¹ and Ωm ≈ 0.30.

  7. Hardware-Oriented Inference Complexity of Kolmogorov-Arnold Networks

    cs.LG 2026-04 conditional novelty 5.0

    Platform-independent formulas for KAN hardware inference complexity (RM, BOP, NABS) are derived for B-spline, GRBF, Chebyshev, and Fourier variants.

  8. Interpretable Clinical Classification with Kolmogorov-Arnold Networks

    cs.LG 2025-09 conditional novelty 5.0

    Logistic KAN and KAAM achieve competitive or superior accuracy on clinical datasets compared to linear, tree, and neural baselines while providing built-in interpretability via symbolic forms and feature-wise decompositions.

  9. P1-KAN: an effective Kolmogorov-Arnold network with application to hydraulic valley optimization

    cs.LG 2024-10 unverdicted novelty 5.0

    P1-KAN introduces a new KAN architecture with theoretical approximation guarantees that outperforms MLPs and prior KAN variants on irregular functions while matching spline KAN accuracy on smooth ones, demonstrated on...

  10. Kolmogorov Arnold networks (KAN) for aerodynamic prediction: a comparison with MLPs and GNNs

    cs.LG 2026-06 unverdicted novelty 4.0

    KANs achieve good but marginally inferior performance to MLPs for airfoil pressure coefficient prediction while being simpler and faster to train, though unstable and hyperparameter-sensitive; GNNs perform best.

  11. A Practitioner's Guide to Kolmogorov-Arnold Networks

    cs.LG 2025-10 accept novelty 3.0

    A systematic review of Kolmogorov-Arnold Networks that maps their relation to Kolmogorov superposition theory, MLPs, and kernels, examines basis-function design choices, summarizes performance advances, and supplies a...

  12. From Handcrafted Features to Functional Edge Learning: Evolution of EEG Seizure Detection Frameworks

    cs.LG 2026-06 unverdicted novelty 2.0

    A review arguing that Kolmogorov-Arnold Networks address key limitations of deep learning models for EEG-based seizure detection through improved interpretability and efficiency.