AI RESEARCH
CausalEvolve: Towards Open-Ended Discovery with Causal Scratchpad
arXiv CS.LG
•
ArXi:2603.14575v1 Announce Type: new Evolve-based agent such as AlphaEvolve is one of the notable successes in using Large Language Models (LLMs) to build AI Scientists. These agents tackle open-ended scientific problems by iteratively improving and evolving programs, leveraging the prior knowledge and reasoning capabilities of LLMs. Despite the success, existing evolve-based agents lack targeted guidance for evolution and effective mechanisms for organizing and utilizing knowledge acquired from past evolutionary experience.