Re4: Scientific Computing Agent with Rewriting, Resolution, Review and Revision

Researchers have developed Re4, a multi-agent AI framework that uses three specialized LLMs (Consultant, Reviewer, and Programmer) working collaboratively to solve scientific computing problems. The system employs a rewriting-resolution-review-revision process that significantly improves bug-free code generation and reduces non-physical solutions in mathematical and scientific reasoning tasks.

• →Re4 framework uses three specialized LLMs in collaborative roles to tackle complex scientific computing problems.
• →The system implements a four-step process: rewriting, resolution, review, and revision for autonomous code generation.
• →Testing on partial differential equations and data-driven analysis shows significant improvement over single-model approaches.
• →The framework reduces non-physical solutions and increases bug-free code generation rates through iterative self-debugging.
• →Code is open-source and available on GitHub, enabling broader research community adoption.