Content
@
0 reply
0 recast
0 reaction
Evan Pappas
@epappas
New architecture for LLMs: bRAG; You ask the LLM something, and the dude acts like it knows it, but won't tell you because you're not cool enough. (yes it's multimodal, and private - works even on 1B params).
0 reply
0 recast
0 reaction
