Watching a language model think before it speaks

nninjahawk1 about 8 hours ago 1 commentsRead Article on blog.nathanlangley.dev

DE version is available. Content is displayed in original English for accuracy.

Discussion (1 Comments)Read Original on HackerNews

mherdelight•about 2 hours ago

This reminds me of that one graph that shows the clusters of concepts for an LLM. Very interesting.