Zero-Shot Learning
Zero-shot learning is when a model performs a task it was not specifically trained for, without being given any examples, relying only on instructions and its general knowledge.
What is Zero-Shot Learning?
Asking an LLM to classify customer reviews as positive or negative with no examples is zero-shot use. Strong zero-shot ability is a hallmark of foundation models.
How it works
The model maps the instruction onto knowledge acquired during pre-training to produce the answer.
← All Artificial Intelligence & Superintelligence definitions · ← Back to the A–Z definitions