AuroraNemoia’s Twitter Archive—№ 646

@zhil_arf we need to poast about how RAG is better than every other method of LLM augmentation and publish a non-academically endorsed paper demonstrating complex reasoning from RAG on a 1B model... asap