Generative AI Decoded: Charts vs. Tables for Data Comprehension

This study delves into how different input formats (charts versus tables) affect a Large Language Model (LLM)'s ability to interpret financial data. The research utilizes several cutting-edge models to assess their performance in both reading comprehension and summarization tasks. It's a fascinating look at how we can optimize prompts for better results.

• •The study used various Generative AI models to compare comprehension of data presented in graphs versus tables.
• •The research used FX (Foreign Exchange) data, input as both graph images and tables.
• •The models were evaluated on their ability to answer reading comprehension questions and provide summaries.