README.md

metadata

title: Static vs Contextual Embeddings
sdk: gradio
python_version: '3.12'
app_file: app.py

Static vs Contextual Embeddings

This Gradio app compares every pair in a word list using static GloVe similarity and contextual token similarity from Alibaba-NLP/gte-modernbert-base.

Enter a pre-context, a comma- or newline-separated word list, and a post-context. The app builds one sentence per word and reports pairwise similarities.