Evaluation
Data contamination
When benchmark questions or answers appear in a model's training data, inflating its score.
Contamination makes a model look better at a test than at the skill the test measures. Newer, regularly refreshed benchmarks are harder to contaminate.
Frequently asked questions
What does Data contamination mean?
When benchmark questions or answers appear in a model's training data, inflating its score. Contamination makes a model look better at a test than at the skill the test measures. Newer, regularly refreshed benchmarks are harder to contaminate.