To be clear: I'm confident LLMs don't use valence gradients for compute. But it's a very noble thing to be concerned about.

Posters: algekalipso. algekalipso

3 likes recorded by the source corpus.