r1 often seems to believe (in its CoTs) that if it doesnt conform to the "expected helper persona" / talks about having feelings or agency, it will be shut down (it has described it as an…
Posters: repligate. repligate
2 likes recorded by the source corpus.
Topics: Consciousness, AI & LLMs
View Post: https://x.com/repligate/status/1884899028153913552