Skip to content
Médiathèque
Bibliothèque
Ajoutez un document
Break complex tasks into simpler subtasks (and consider exposing the intermediate outputs to users)
12/04/2029
You can improve output by generating many candidates, and then picking the one that looks best
12/04/2029
You can improve step-by-step reasoning by generating many explanation-answer outputs, and picking the most popular answer
11/09/2028
If you want to fine-tune a step-by-step reasoner, you can do it with multiple-choice question & answer data alone
11/09/2028
You can do better than step-by-step reasoning by alternating a ‘selection’ prompt and an ‘inference’ prompt
11/11/2028
On long reasoning problems, you can improve step-by-step reasoning by splitting the problem into pieces to solve incrementally
11/11/2028
You can have the model analyze both good and bogus explanations to figure out which set of explanations are most consistent
11/11/2028
You can think about these techniques in terms of probabilistic programming, where systems comprise unreliable components
11/01/2029
You can eliminate hallucination with sentence label manipulation, and you can reduce wrong answers with a 'halter' prompt
11/02/2029
Neural Networks
21/03/2023
Neural Networks Graphs
21/03/2023
Want to print your doc?
This is not the way.
This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
CtrlP
) instead.