Concept
Feature-based
Feature-based refers to using a pre-trained model as a fixed source of representations, rather than continuing to train its weights on your task. You run your inputs through the frozen model, extract the activation vectors it produces, and feed those vectors as features into…
The rest of “Feature-based” 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→