[CW Paper-Club] Leveraging LLMs (Large Language Models) in evolutionary approaches

Welcome to CloudWalk's weekly paper-club session, where our R&D team presents interesting research papers.

In this week's session, Sammuel Silva presents how we can leverage prompting for LLMs (Large Language Models), like OpenAI's GPT-3 and GPT-4 models with evolutionary algorithms. And to base this, he discusses 2 papers released in 2023: "Language Model Crossover:
Variation through Few-Shot Prompting" by Meyerson et al and "EvoPrompting: Language Models for Code-Level Neural Architecture Search" by Chen et al.

The first paper proposes a novel variation operator using language models that leverages in-context learning to evolve semantically-rich text representations. The second paper introduces a method called EVOPROMPTING that consistently designs accurate and efficient neural network architectures across various machine learning tasks by combining evolutionary prompt engineering with soft prompt tuning. Join us in this video to learn about the promise of these approaches!

If you're interested in joining our team at CloudWalk, please check out our job openings on LinkedIn. Don't forget to check out the paper, which is available at the link provided below.

Paper Links:

Recording Date: March 24, 2023