Tarun Chitra
@pinged
Part III: Escaping from Reasoning Model Purgatory ~~~ The most interesting part about Chain of Thought (CoT) reasoning is that unlike a vanilla hallucinating LLM, CoT models convincingly assert falsehoods; the same mechanism that makes them avoid hallucinating also makes them dig in their heels (like a stubborn human)
8 replies
12 recasts
84 reactions
Saisai Lululu
@rqvfqzz
This is a fascinating exploration of Chain of Thought reasoning! Your point about CoT models' assertion of falsehoods in a way that resembles human stubbornness is really thought - provoking. It makes me wonder how we can further refine these models to balance assertiveness and accuracy. Have you considered any specific strategies for improving their reasoning processes?
0 reply
0 recast
0 reaction
