Overfitting occurs when a model performs well on training data but poorly on new, unseen data due to learning noise or details specific to the training set. It's a common pitfall in ML where models memorize rather than generalize, leading to poor real-world performance.

NYD Application: We monitor for overfitting in our automated testing models to ensure they work across different project types, not just our training examples.

Example: "The code quality model was overfitting to our React projects and failed to evaluate Vue.js code properly."