Can LLMs Generate Visualizations with Dataless Prompts?

TL;DR

This paper explores whether large language models like GPT-3 and GPT-4 can generate accurate data visualizations from simple, dataless prompts, comparing their outputs to expert-created cheat sheets.

Contribution

It is the first study to evaluate LLMs' ability to produce visualizations without data prompts, highlighting their potential and limitations in visualization tasks.

Findings

GPT-3 and GPT-4 can generate relevant visualizations from dataless prompts.

Models' outputs are comparable to expert-created cheat sheets in some cases.

The study identifies key challenges and future directions for LLM-based visualization generation.

Abstract

Recent advancements in large language models have revolutionized information access, as these models harness data available on the web to address complex queries, becoming the preferred information source for many users. In certain cases, queries are about publicly available data, which can be effectively answered with data visualizations. In this paper, we investigate the ability of large language models to provide accurate data and relevant visualizations in response to such queries. Specifically, we investigate the ability of GPT-3 and GPT-4 to generate visualizations with dataless prompts, where no data accompanies the query. We evaluate the results of the models by comparing them to visualization cheat sheets created by visualization experts.