Difference between revisions of "Softmax function"

From wikieduonline
Jump to navigation Jump to search
Tags: Mobile web edit, Mobile edit
Tags: Mobile web edit, Mobile edit
 
Line 5: Line 5:
 
A smooth approximation to the [[maximum]] function
 
A smooth approximation to the [[maximum]] function
  
 +
 +
* [[Cross-entropy]]
  
 
{{NN}}
 
{{NN}}

Latest revision as of 06:48, 10 September 2025

wikipedia:Softmax function softargmax

The softmax function is often used as the last activation function of a neural network to normalize the output of a network to a probability distribution over predicted output classes.

A smooth approximation to the maximum function


Artificial neural networks, Neuronal network (NN), CNN, RNN, Micrograd, NPU, ConvNet, AlexNet, GoogLeNet, Apache MXNet, Neural architecture search, DAG, Feedforward neural network, NeurIPS, Feature Pyramid Network, TPU, NPU, Apple Neural Engine (ANE), LLM, TFLOPS, Softmax function, Dilution (neural networks), AlphaGo

Advertising: