---
title: Kolmogorov-Arnold-Networks vs ANN
url: https://www.ml-quant.com/papers/ssrn/5072413/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: SSRN 5072413
source_url: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5072413
featured: 2025-01-01
citations: unknown
topic: ML & AI Methods
---


# Kolmogorov-Arnold-Networks vs ANN

The paper mathematically compares Kolmogov-Arnold-Networks and traditional Artificial Neural Networks, showing a balance between expressivity and computational complexity.

- Source: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5072413
- Identifier: SSRN 5072413
- Released: 2024-12-26
- First featured: Quant Letter No. 80 (2025-01-01): https://www.ml-quant.com/issues/2025-01-01/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: ML & AI Methods

## Related

- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://www.ml-quant.com/papers/arxiv/2312.00752/): Sequence Modeling: Mamba, a neural network architecture that doesn't use attention or MLP blocks, provides faster inference and better performance in language, audio, and genomics than Transformers.
- [Graph Mamba: Towards Learning on Graphs with State Space Models](https://www.ml-quant.com/papers/arxiv/2402.08678/): Graph Mamba Networks, a new type of Graph Neural Networks, have been introduced, which achieve excellent performance in various benchmark datasets despite lower computational cost.
- [Edge Directionality Improves Learning on Heterophilic Graphs](https://www.ml-quant.com/papers/arxiv/2305.10498/): The study presents Directed Graph Neural Network (Dir-GNN), a new deep learning framework for directed graphs that surpasses traditional models in heterophilic benchmarks.
- [Tensor Programs VI: Feature Learning in Infinite-Depth Neural Networks](https://www.ml-quant.com/papers/arxiv/2310.02244/): Deep Residual Network Feature Learning: The research explores depthwise parametrizations in deep residual networks, pinpointing Depth-$\mu$P as the best parametrization for maximizing feature learning and diversity, but notes its limitations in deeper networks.
- [Modular Duality in Deep Learning](https://www.ml-quant.com/papers/arxiv/2410.21265/): The article presents a new theory of modular dualization for general neural networks, providing a theoretical basis for fast and scalable training algorithms, potentially leading to a new generation of optimizers for neural architectures.
- [IMAGINE: An 8-to-1b 22nm FD-SOI Compute-In-Memory CNN Accelerator With an End-to-End Analog Charge-Based 0.15-8POPS/W Macro Featuring Distribution-Aware Data Reshaping](https://www.ml-quant.com/papers/arxiv/2412.19750/): The paper introduces IMAGINE, a compute-in-memory SRAM for processing convolutional neural networks, which offers high energy efficiency and competitive accuracies on MNIST and CIFAR-10.