Learn Before
Example

Token Classification in Discriminative Training

Token classification is a discriminative pre-training approach used for encoder-only models. In this method, the model receives an input sequence and performs a classification task for each individual token. For example, given the input [C] The kitten is chasing the ball ., the model outputs a supervision signal or probability distribution for every token, such as Pr(The)\Pr(\cdot|\textrm{The}) and Pr(kitten)\Pr(\cdot|\textrm{kitten}).

0

1

Updated 2026-04-17

Contributors are:

Who are from:

Tags

Foundations of Large Language Models

Ch.1 Pre-training - Foundations of Large Language Models

Foundations of Large Language Models Course

Computing Sciences