|
|
|
|
|
|
by kgeist
127 days ago
|
|
|
LLMs often already "know" the answer starting from the first output token and then emulate "reasoning" so that it appeared as if it came to the conclusion through logic. There's a bunch of papers on this topic. At least it used to be the case a few months ago, not sure about the current SOTA models. |
|
|