|
|
|
|
|
|
by usgroup
784 days ago
|
|
|
A consequence of being an auto regressive model is not being able to plan token output. I think the author’s example is one of the many corollaries. You could prompt the LLM differently , for example to write a Python program that does the random part, and then act on its output. |
|
|
Generating independent simple random variables requires zero planning by definition, because they are independent. And base auto-regressive models do it fine.