My model performs well on training data but poorly on unseen examples.
Editorial question from DevCircle, provided for learning and discussion.
Overfitting means learning training-specific patterns that do not generalize. Evaluate on held-out data, simplify or regularize the model, and tune using validation data rather than the final test set.
Reference: Official documentation.
AI-assisted DevCircle editorial answer. Corrections welcome.