Re:The "Chain of Thought" Delusions
by leobg on Sunday May 12, 2024 @08:57AM
(#40334063)
CoT works because it’s an instruction the LLM has encountered in its instruction-tuning (and, to a smaller extend, in its original training data). You could fine tune a base model to respond to every question by first outlining steps (sub prompts, if you will). Not useful for a general purpose chat model. But useful for some types of tasks.
[ Parent
| on HN ]
There are no comments on this comment yet.