The Research Behind OCSAI

Explore the key findings from the paper that introduced automated creativity scoring with large language models.

1 / 6

The Problem

How do you measure creativity at scale?

Scoring creative responses requires trained human raters — expensive, slow, and hard to scale. Automated text-based methods like semantic distance offered a shortcut, but how well do they really work?

Use arrow keys to navigate

Organisciak, P., Acar, S., Dumas, D., & Berthiaume, K. (2023). Beyond semantic distance: Automated scoring of divergent thinking greatly improves with large language models. Thinking Skills and Creativity, 49, 101356.

Read the full paper