Javier Augusto Galvis Chacon

Learning Differential Evolution Mutation Strategies via Performance-Driven Large Language Models

Experiments on BBOB show the feedback loop yields mutation strategies that outperform classical DE operators and LLM-generated variants without feedback.