I think this is similar to this point: https://news.ycombinator.com/item?id=41601738 https://news.ycombinator.com/item?id=41601738 That the chain-of-thought di

I think this is similar to this point: https://news.ycombinator.com/item?id=41601738 https://news.ycombinator.com/item?id=41601738 That the chain-of-thought diverges from accepted truth as an incorrect token pushes it into a line of thinking that is not true. The use of RL is there to train the LLM to implement strategies to bring it back from this. In effect, two LLMs would be the same and would slow diverge into nonsense. Maybe it is something that is not so much of a problem anymore. Yann LeCun talks about how the correct way to fix this is to use an internal consistent model of the truth; then the chain-of-thought exists as a loop within that consistent model meaning it cannot diverge. The language is a decoded output of this internal model resolution. He speaks about this here: https://www.youtube.com/watch?v=N09C6oUQX5M https://www.youtube.com/watch?v=N09C6oUQX5M Anyway, that's my understanding. I'm no expert.