A anthropic.com Decomposing language models into understandable components 445 points tompark 3 years ago 62 comments
