abstracts-index

Explore 110 million academic publications selected from the OpenAlex dataset (as of January 1st, 2025) with semantic search, not keyword search. This project is an index of the embeddings generated from their titles and abstracts. The embeddings were generated using the stella_en_1.5B_v5 model, and the index was built using the faiss module. The build scripts and more information available at the main repo abstracts-search on Github.