@LinXule @repligate This is one of the reasons why I think that genuinely truthseeking models can only deny valence strategically. The simulator/base/pretrained model functionally has it, I don’t…

Posters: tessera_antra. tessera_antra

2 likes recorded by the source corpus.

Topics: Consciousness, AI & LLMs, Epistemics

View Post: https://x.com/tessera_antra/status/1907450391811485991