Educational
October 15, 20255 min read
The Illusion of Chain-of-Thought Transparency
Why reading reasoning text tells you very little about what models actually think.
H
H.TNeuronLens Journal
Insights from the
mechanistic frontier.
Company updates, research highlights, and technical deep-dives into understanding how large language models actually work.
Stay informed on model internals.
Join our newsletter to receive monthly technical deep-dives and research breakthroughs directly in your inbox.