SUPN: Shallow Universal Polynomial Networks for Efficient Function Approximation
cs.LG, cs.NA, math.NA
Submitted: 2025-11-26
Updated: 2026-10-15
Comments: 47 pages including appendices
Journal ref: Trans. Mach. Learn. Res. (2026), ISSN 2835-8856
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Leveraging KANs for Expedient Training of Multichannel MLPs via Preconditioning and Geometric Refinement
- Growing axons: greedy learning of neural networks with application to function approximation
- Universal Approximation with Deep Narrow Networks
- Fourier Neural Operator for Parametric Partial Differential Equations
- Kernel Neural Operators (KNOs) for Scalable, Memory-efficient, Geometrically-flexible Operator Learning
- Collapse of Deep and Narrow Neural Nets
- On First-Order Meta-Learning Algorithms
- Chebyshev Polynomial-Based Kolmogorov-Arnold Networks: An Efficient Architecture for Nonlinear Function Approximation
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks