Learn Before
Applying the Text-Pair Classification Pipeline
A developer is building a model to determine if a user review contradicts a product description. The model formats the input as a single sequence, processes it through an encoder, and produces a sequence of contextualized hidden state vectors, one for each input token. Given the final hidden state vectors shown below, which specific vector should be passed to the final prediction network to classify the relationship between the two texts? Justify your answer based on the standard architecture for this task.
0
1
Tags
Ch.1 Pre-training - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Ch.2 Generative Models - Foundations of Large Language Models
Application in Bloom's Taxonomy
Cognitive Psychology
Psychology
Social Science
Empirical Science
Science
Related
Schematic Example of a Sentence-Pair Classification Pipeline
In a text-pair classification pipeline, two texts are formatted into a single input sequence, which includes a special token at the beginning designated for classification. After a Transformer encoder processes this sequence and generates a contextualized hidden state for every token, a single vector must be selected to represent the entire text pair for the final prediction. Which of the following best explains the standard method for selecting this vector and the reasoning behind it?
A language model is tasked with determining if two sentences are semantically equivalent. Arrange the following steps to correctly represent the end-to-end computational pipeline, from preparing the input to generating the final prediction.
Applying the Text-Pair Classification Pipeline