Expressive power of ReLU and step networks under floating-point operations

DOI10.1016/J.NEUNET.2024.106297MaRDI QIDQ6543653

Sejun Park, Geonho Hwang, Yeachan Park, Won-Yeol Lee

Publication date: 24 May 2024

Published in: Neural Networks (Search for Journal in Brave)

zbMATH Keywords

neural networksfloating-point arithmeticuniversal approximationmemorization

Mathematics Subject Classification ID

Numerical analysis (65-XX)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Sharp error bounds for complex floating-point inversion
• On the capabilities of multilayer perceptrons
• Multilayer feedforward networks are universal approximators
• Stupid is as stupid does: taking the square root of the square of a floating-point number
• Exploiting Structure in Floating-Point Arithmetic
• Elementary Functions
• On relative errors of floating-point operations: Optimal bounds and applications
• Handbook of Floating-Point Arithmetic
• Memory Capacity of Neural Networks with Threshold and Rectified Linear Unit Activations
• Approximation by superpositions of a sigmoidal function
• Floating-point arithmetic

This page was built for publication: Expressive power of ReLU and step networks under floating-point operations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6543653)