juniorGradient Descent
What is the gradient in Gradient Descent?
Updated May 16, 2026
Short answer
The gradient is a vector of partial derivatives indicating direction of steepest increase.
Deep explanation
The gradient shows how much the function changes with respect to each parameter. Gradient Descent uses the negative gradient to move toward minimum values.
Real-world example
Adjusting weights in neural networks based on error sensitivity.
Common mistakes
- Ignoring that gradient is vector-valued in multivariate functions.
Follow-up questions
- What is partial derivative?
- What is gradient direction?