Large Language Models Are Human-Level Prompt Engineers abs: arxiv.org/abs/2211.01910 project page: sites.google.com/view/automatic…

Can large language models write prompts…for themselves? Yes, at a human-level (!) if they are given the ability to experiment and see what works. arxiv.org/abs/2211.01910 with Yongchao Zhou, Andrei Muresanu, Ziwen, Silviu Pitis, Harris Chan, and Jimmy Ba (1/7)

Searching for the right prompt? APE makes it easy to find a prompt that works for your language modeling task. Try our official release: Colab: tiny.cc/ape_demo GUI: tiny.cc/ape_ui GitHub: github.com/keirp/automati…

Large Language Models Are Human-Level Prompt Engineers "Inspired by the human approach to prompt engineering ... ... we achieve better or comparable performance to the instructions generated by human annotators on 19/24 tasks" arxiv: arxiv.org/abs/2211.01910

“Let’s work this out in a step by step way to be sure we have the right answer” is now on the leaderboard!

Try out our Automatic Prompt Engineering (APE) Studio on your use cases. Tweet us your amazing prompts! GUI: tiny.cc/ape_ui

Getting good answers from an artificial intelligence can be tricky, and now researchers have found that an AI is better at asking questions of another AI than people are newscientist.com/article/234651…

Neat result from applying automated prompt generation on translation. Referencing "Google Translate" in the prompt improves the performance of the model.

One of my favorite results in 2022 was that it's not enough to just think step by step. You must also make sure to get the right answer :D sites.google.com/view/automatic… (actually a nice insight into a psychology of a GPT; it pays to condition on a high reward)