Best Practices and Instruction Methods for Prompt Generation in ChatGPT, led by Joelle Mornini (she/her/hers) and Alicia Lillich with the ORS Division of Library Services, National Institutes of Health Library, will be held on February 4, 2025, 11:00 AM-12:00 PM PST/1:00-2:00 PM CST/2:00-3:00 PM EST.
This session equips participants with essential knowledge and skills for effective interactions with Large Language Models (LLMs), such as ChatGPT. Participants will learn about prompt patterns to optimize LLMs’ capabilities and explore how librarians can plan effective training to promote AI adoption through instruction and services.
By the end of this session, participants will be able to:
– Define LLMs, prompt patterns, and prompt engineering
– Apply prompt patterns to improve generated output from LLMs
– Teach prompt engineering skills to library patrons in a biomedical library setting
This session provides an overview on Language Models (LLMs), ChatGPT, prompt engineering, and prompt patterns, including definitions, descriptions of potential uses for LLMs in the biomedical research field, a live demonstration of ChatGPT that introduces the tool’s user interface, other examples of LLMs beyond ChatGPT, and ethical considerations to keep in mind when using LLMs. The class then introduces the article “A Prompt Pattern Catalog to Enhance Prompt Engineering with ChatGPT” (https://arxiv.org/abs/2302.11382v1) and provides descriptions and examples of prompt patterns to improve output from LLMs. The example prompts are built around a scenario of planning a class for college-age interns about identification of predatory journals.