{"id":173837,"date":"2023-10-09T11:24:29","date_gmt":"2023-10-09T16:24:29","guid":{"rendered":"https:\/\/lifeboat.com\/blog\/2023\/10\/promptbreeder-self-referential-self-improvement-via-prompt-evolution-paper-explained"},"modified":"2023-10-09T11:24:29","modified_gmt":"2023-10-09T16:24:29","slug":"promptbreeder-self-referential-self-improvement-via-prompt-evolution-paper-explained","status":"publish","type":"post","link":"https:\/\/lifeboat.com\/blog\/2023\/10\/promptbreeder-self-referential-self-improvement-via-prompt-evolution-paper-explained","title":{"rendered":"Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution (Paper Explained)"},"content":{"rendered":"<p><\/p>\n<p><iframe style=\"display: block; margin: 0 auto; width: 100%; aspect-ratio: 4\/3; object-fit: contain;\" src=\"https:\/\/www.youtube.com\/embed\/tkX0EfNl4Fc?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; encrypted-media; gyroscope;\n   picture-in-picture\" allowfullscreen><\/iframe><\/p>\n<p>#evolution.<\/p>\n<p>Promptbreeder is a self-improving self-referential system for automated prompt engineering. Give it a task description and a dataset, and it will automatically come up with appropriate prompts for the task. This is achieved by an evolutionary algorithm where not only the prompts, but also the mutation-prompts are improved over time in a population-based, diversity-focused approach.<\/p>\n<p>OUTLINE:<br \/> 0:00 \u2014 Introduction.<br \/> 2:10 \u2014 From manual to automated prompt engineering.<br \/> 10:40 \u2014 How does Promptbreeder work?<br \/> 21:30 \u2014 Mutation operators.<br \/> 36:00 \u2014 Experimental Results.<br \/> 38:05 \u2014 A walk through the appendix.<\/p>\n<p>Paper: <a href=\"https:\/\/arxiv.org\/abs\/2309\">https:\/\/arxiv.org\/abs\/2309<\/a>.<\/p>\n<p>Abstract:<br \/> Popular prompt strategies like Chain-of-Thought Prompting can dramatically improve the reasoning abilities of Large Language Models (LLMs) in various domains. However, such hand-crafted prompt-strategies are often sub-optimal. In this paper, we present Promptbreeder, a general-purpose self-referential self-improvement mechanism that evolves and adapts prompts for a given domain. Driven by an LLM, Promptbreeder mutates a population of task-prompts, and subsequently evaluates them for fitness on a training set. Crucially, the mutation of these task-prompts is governed by mutation-prompts that the LLM generates and improves throughout evolution in a self-referential way. That is, Promptbreeder is not just improving task-prompts, but it is also improving the mutationprompts that improve these task-prompts. Promptbreeder outperforms state-of-the-art prompt strategies such as Chain-of-Thought and Plan-and-Solve Prompting on commonly used arithmetic and commonsense reasoning benchmarks. Furthermore, Promptbreeder is able to evolve intricate task-prompts for the challenging problem of hate speech classification.<\/p>\n<p>Authors: Chrisantha Fernando, Dylan Banarse, Henryk Michalewski, Simon Osindero, Tim Rockt\u00e4schel.<\/p>\n<div class=\"more-link-wrapper\"> <a class=\"more-link\" href=\"https:\/\/lifeboat.com\/blog\/2023\/10\/promptbreeder-self-referential-self-improvement-via-prompt-evolution-paper-explained\">Continue reading \u201cPromptbreeder: Self-Referential Self-Improvement Via Prompt Evolution (Paper Explained)\u201d | &gt;<\/a><\/div>\n","protected":false},"excerpt":{"rendered":"<p>#evolution. Promptbreeder is a self-improving self-referential system for automated prompt engineering. Give it a task description and a dataset, and it will automatically come up with appropriate prompts for the task. This is achieved by an evolutionary algorithm where not only the prompts, but also the mutation-prompts are improved over time in a population-based, diversity-focused [\u2026]<\/p>\n","protected":false},"author":556,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[38,385,41],"tags":[],"class_list":["post-173837","post","type-post","status-publish","format-standard","hentry","category-engineering","category-evolution","category-information-science"],"_links":{"self":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/173837","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/users\/556"}],"replies":[{"embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/comments?post=173837"}],"version-history":[{"count":0,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/173837\/revisions"}],"wp:attachment":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/media?parent=173837"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/categories?post=173837"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/tags?post=173837"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}