S shyam.blog Beyond self-attention: How a small language model predicts the next token 474 points tplrbv 2 years ago 85 comments
