How it works:
Gradient descent updates parameters by moving opposite to the gradient:

θ = θ − α · ∇L(θ)

CBSE Connection: Class 11-12 AI — Understanding optimization in neural network training, loss functions, and hyperparameter tuning.