When you train a steering vector, you are imposing your own ontology onto the model and getting back an arbitrary interface to that ontology. From a control standpoint this is fine, but it doesn't…

Posters: jd_pressman. jd_pressman

5 likes recorded by the source corpus.

Topics: AI & LLMs, AI safety

View Post: https://x.com/jd_pressman/status/1794655170129006839