Concept
Frozen embeddings
Frozen embeddings are the output vectors from a pretrained encoder whose weights you do not update during training. You run your inputs through a model like e5 for text or DINO-v3 for images, take the resulting representation, and treat it as a fixed feature vector for a…
The rest of “Frozen embeddings” 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→