@plain-english-ai

11 presentations

Subscribe
LatestPopular
5:34

RAG has an AI look up passages before it answers

@plain-english-ai · 2 weeks ago

5:24

Knowledge distillation copies a teacher model’s wrong guesses

@plain-english-ai · 2 weeks ago

5:46

LoRA fine-tunes GPT-3 by training 4.7 million new numbers

@plain-english-ai · 2 weeks ago

5:49

RLHF taught GPT-3 to reply using rankings from about 40 people

@plain-english-ai · 2 weeks ago

5:57

GPT-3 learned to write by guessing the next piece 300 billion times

@plain-english-ai · 2 weeks ago

5:36

ChatGPT uses attention to work out what “its” points to

@plain-english-ai · 2 weeks ago

5:54

Training GPT-3 cost 900 billion times one piece of an answer

@plain-english-ai · 2 weeks ago

5:09

Word2Vec puts each word on a map of company

@plain-english-ai · 1 month ago

5:50

What a context window actually holds

@plain-english-ai · 1 month ago

5:18

Why AI reads in tokens, not words

@plain-english-ai · 1 month ago

3:17

Why chatbots make things up — and why more data doesn't fix it

@plain-english-ai · 1 month ago