- author = "Chrisantha Fernando and Dylan Banarse and Henryk Michalewski and Simon Osindero and Tim Rocktachel",
- title = "Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution",
- howpublished = "arXiv 2309.16797",
- year = "2023",
- month = "28 " # sep,
- keywords = "genetic algorithms, genetic programming, ANN, LLM",
- eprint = "2309.16797",
- archiveprefix = "arXiv",
-
URL = "
https://arxiv.org/pdf/2309.16797",
- size = "64 pages",
- abstract = "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. 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 mutation prompts 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.",
-
notes = "'We then run a standard binary tournament genetic
algorithm'
Google DeepMind",
