Z
4 terms
Showing all terms starting with Z
Zero-Shot Learning
The ability of an AI model to perform a task without having seen any labelled examples for that task during training or in the prompt.
Z-score Normalisation
A data preprocessing technique that scales features to have zero mean and unit variance, improving training stability in many ML models.
Zero-Knowledge Proof (AI)
A cryptographic technique increasingly used in AI to verify model outputs or data without revealing the underlying private information.
ZeRO (Optimiser)
Zero Redundancy Optimiser - a memory optimisation strategy for training very large models across multiple GPUs by eliminating parameter redundancy.
