underfitting and overfitting
Used in artificial intelligence and machine learning, underfitting and overfitting are the two primary LLM failure modes when a model tries to learn the relationship between input data and target outcomes.
Underfitting is when a model is too simple or undertrained to capture the underlying patterns in the data, resulting in a high bias, with poor accuracy.
Overfitting is when an AI memorizes the training data too closely instead of learning general rules, making it bad at new examples. This results in low error on the training set (high accuracy), but high error on new, validation, or test data (poor accuracy).
The ultimate goal of machine learning is generalization: Making accurate predictions on new, unseen data. When a model learns too little or too much, we get underfitting or overfitting.
See also : WITFITS if I tell you, I'll have to kill you artificial intelligence machine learning singularity
NetLingo Classification: Net Technology
Edit Word