Independent Variable

Also known as: Predictor, Feature, Input Variable, Explanatory Variable

An independent variable is an input used to explain or predict a dependent variable. In ML it is usually called a feature or predictor: tabular columns, tokens, pixels, or embeddings that the model conditions on.

Why it matters

  • Feature choice and quality dominate classical ML performance
  • Collinear or leaked features distort interpretation and metrics
  • In LLMs, “features” are largely learned representations; the explicit inputs are tokens and context

How it shows up

Regression write-ups still say $X$ for independents and $y$ for the target. Feature engineering builds $X$; supervised learning fits $f(X) \approx y$. Watch for multicollinearity when interpreting coefficients, and for train–serve skew when production features drift from training.

In generative settings the conditioning context (prompt, retrieved chunks, tool results) plays the same role: inputs the model conditions on to produce the next outputs.

← Back to AI Glossary