AI glossary · Machine learning
Loss function
A loss function measures how wrong an AI model’s predictions are during training. Training tries to make this number as small as possible.
Also called cost function, objective function
In more detail
After each guess, the loss function gives a score: high when the prediction is far from the right answer, low when it is close. Gradient descent then adjusts the model to reduce the score.
Choosing what to measure matters: a model gets good at whatever its loss function rewards.
Related terms
Gradient descentGradient descent is the method most AI models use to learn: it repeatedly nudges the model’s settings in the direction that reduces its errors.BackpropagationBackpropagation is how a neural network learns from mistakes: it works backwards from the error to find how much each connection caused it, then adjusts them.ParametersParameters are the internal numbers an AI model learns during training. A model’s size, such as “70 billion parameters”, counts them.
Browse related tools
AI tools listed on BroUseAI in these categories.
