Reinforcing Agentic Creativity in Scientific Ideation with Night Science

2026-09-28

AI Summary

This study addresses the homogenization of scientific ideation caused by the low-entropy bias of large language models by proposing the "AI Night Scientist" framework. Grounded in cognitive science, this framework models creativity as a tripartite capacity encompassing action, process, and outcome. By integrating an agent-based architecture with the Group Relative Policy Optimization (GRPO) reinforcement learning algorithm, it drives models beyond conventional reasoning through semantic guidance rather than mere temperature scaling, thereby facilitating diverse scientific exploration. Experimental results demonstrate that the proposed approach increases research proposal diversity by 27.8% and contribution type variety by 14.9%, while improving predicted citation impact by 32 percentage points and originality scores by 66.2 points, significantly enhancing AI-driven scientific creativity.

Abstract

Large language models (LLMs) excel at structured, verifiable tasks, but their low-entropy bias can produce homogeneous and predictable outputs, limiting their utility for open-ended scientific ideation. Effective discovery, however, spans a broader creative spectrum: from structured day science to loosely structured, serendipitous night science that reaches ideas beyond those typically considered. We introduce AI Night-Scientist, an agentic framework that uses reinforcement learning to teach models when and how to depart from predictable reasoning. Grounded in cognitive science, we model creativity along three axes: action (what to do and how creatively), process (when to explore versus exploit), and outcome (the novelty and usefulness of the resulting idea). We use these axes to train models with GRPO, exposing them to varying degrees and forms of creativity throughout training. This produces substantially more diverse scientific proposals, expanding the range of research directions by 27.8% and contribution types by 14.9% over the base model. It also improves predicted citation impact by up to 32.0 percentage points and originality by 66.2 points. These gains cannot be reproduced by simply increasing decoding temperature; instead, we find that semantic guidance specifying what kind of creativity to pursue is critical. Overall, our results suggest that creativity is a learnable, multi-level ability that can be shaped to help researchers reach ideas beyond those typically explored by LLMs.

Problem

• Scientific Ideation
• Large Language Models
• Agentic Creativity
• Low-entropy Bias
• Night Science

Innovation

• Agentic Framework
• Reinforcement Learning
• GRPO
• Scientific Ideation
• Night Science