Hacker News new | ask | show | jobs
by charcircuit 1009 days ago
His point is that you do not have to rely on associative being true in order to run inference on a LLM.