What is 'Zero-Shot Learning'?

What is 'Zero-Shot Learning'?

Approach: Describe a model's ability to perform a task it was not explicitly trained for.

A machine learning setup where a model can recognize and classify items from classes it has never seen during training, typically by leveraging semantic relationships or high-level descriptions.

Follow-up: How do Large Language Models (LLMs) achieve zero-shot performance?

Key concepts: embeddings, semantics, transfer learning.