data leakage
AI, ML & GenAI interview questions tagged data leakage, across every topic.
2 questions · 1 unlocked for you
Concepts behind "data leakage"
The curriculum that explains the ideas these questions test.
Foundational
Overfitting and RegularizationOverfitting is when a model learns the training data's noise rather than its signal, scoring well in training but failing on new data. You prevent it with more data, regularization (L1/L2, dropout, early stopping), simpler models, and data augmentation, and you detect it with a proper held-out validation set. The deeper trap is data leakage, which yields fake great offline numbers that collapse in production. AI, ML, and GenAI engineer interviews probe it because shipping an overfit or leaky model is one of the most common, expensive ML mistakes.📊 Evaluation & ML Foundations
Core
Benchmarks and Their LimitsPublic benchmarks like MMLU offer a shared yardstick, but they saturate, leak into training corpora, and stop tracking real ability once labs optimize for them. Contamination (test items in the training data) and Goodhart's law (a measure that becomes a target stops measuring) are why a high leaderboard score can mean nothing on your workload. AI, ML, and GenAI engineer interviews probe this to see whether you trust a number or build a private eval set on your own distribution.📊 Evaluation & ML FoundationsSign in