Hacker News

Favorites Setup
Comment by PaulHoule | original | What thousands of conversations look like inside a language model
[−]PaulHoule · 2026-11-15 Sun 15:34 UTC · link
Length-dependence has been a thing in vector search since bag-of-words!

https://en.wikipedia.org/wiki/Okapi_BM25

was one of very few useful results of the first decade of TREC research... And the main thing it addresses is balancing various factors which could cause a full text search engine to privilege either long or short documents.