ExpEvolve: Structuring and Reusing Experience for Open-Ended Discovery

Key points

• Evolutionary coding agents such as AlphaEvolve are one of the notable successes in using Large Language Models (LLMs) to build AI Scientists.1
• These agents tackle open-ended scientific problems by iteratively improving and evolving programs, leveraging the prior knowledge and reasoning capabilities of LLMs. Despite their success, mutations over programs are mostly low-level upon surface heuristics, without a persistent unit of control.1
• We characterize theoretical conditions on when discovery with idea factors improves sample efficiency.1
• Analysis shows that ExpEvolve can discover idea-level factors to provide better exploration and exploitation for evolutionary discovery.1

Sources (1)

• [1]ExpEvolve: Structuring and Reusing Experience for Open-Ended Discovery
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 21, 07:00 AM

  In this work, we propose ExpEvolve, which injects idea-level extraction and control into evolution: it discovers and leverages the high-level factors from programs that explain scoring differences among different programs.

  Evolutionary coding agents such as AlphaEvolve are one of the notable successes in using Large Language Models (LLMs) to build AI Scientists.

Extractive summary: sentences quoted from the sources.