Show HN: LLM Thought Visualization
https://github.com/ninjahawk/Subtext
ninjahawk1 · 2 days ago
3 comments
https://github.com/ninjahawk/Subtext
ninjahawk1 · 2 days ago
3 comments
penpendian · 2 days ago
amazing
chonghaoju · 2 days ago
The moment that grabs me is that the verdict is internally settled several tokens before "not correct" gets verbalized.
What's the per-token cost of reading the lens at nine depths?
sageserpent · 2 days ago
Thank you for making me aware of the concept of "J-space". Just read the Anthropic post on this, fascinating...
ninjahawk1 · 2 days ago
I was blown away when I read the article yesterday, mech interp has been a black box for so long but it seems we’re finally making advances