Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

average intra-sentence similarity

Average intra-sentence similarity is a natural language processing metric that measures the extent to which the contextualized vector representations of individual words in a sentence resemble the overall sentence representation, averaged across a collection of sentences. For a given sentence, intra-sentence similarity is typically calculated by taking the average cosine similarity between each token embedding and the mean embedding of all words in that sentence. When evaluated across multiple sentences, this measure provides insight into the geometric behavior of language representation models, helping to quantify how much word representations absorb context from their surrounding sentence versus preserving distinct lexical identities across different layers of a model.

1 item

How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT-2 Embeddings

How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT-2 Embeddings

Kawin Ethayarajh

OrganizationsStanford University

Why you should read this

Reveals the geometric behavior of BERT, ELMo, and GPT-2 embeddings across layers, proving that upper layers generate more context-specific representations while static word vectors account for less than five percent of their variance.

Replacing static word embeddings with contextualized word representations has yielded significant improvements on many NLP tasks. However, just how contextual are the contextualized representations produced by models such as ELMo and BERT? Are there infinitely many context-specific representations for each word, or are words essentially assigned one of a finite number of word-sense representations? For one, we find that the contextualized representations of all words are not isotropic in any layer of the contextualizing model. While representations of the same word in different contexts still have a greater cosine similarity than those of two different words, this self-similarity is much lower in upper layers. This suggests that upper layers of contextualizing models produce more context-specific representations, much like how upper layers of LSTMs produce more task-specific representations. In all layers of ELMo, BERT, and GPT-2, on average, less than 5% of the variance in a word's contextualized representations can be explained by a static embedding for that word, providing some justification for the success of contextualized representations.

Added

2026-09-25