Concept
Zero-shot transfer
Zero-shot transfer means applying a trained model to a task or set of labels it never saw during training, without any task-specific fine-tuning or gradient updates. The "zero" refers to zero training examples for the new task; you specify what you want at inference time…
The rest of “Zero-shot transfer” is a premium feature: every concept in the library gets a precise, practitioner-focused write-up like this one, cross-linked straight from the paper summaries that use it.
Log in to unlock→