Researchers have developed AgenticEvolve, a novel framework for evolutionary program search that utilizes persistent large language model agents. These agents are designed to autonomously create and execute experiments on candidate programs between evaluation stages. Comparative analysis shows AgenticEvolve outperforms existing methods like No-analysis and OpenEvolve, achieving superior program quality with fewer evaluations and comparable costs on heuristic design tasks and equation discovery problems. AI
IMPACT Enhances evolutionary algorithms by enabling LLM agents to autonomously design and run experiments, potentially accelerating discovery in fields like heuristic design and equation discovery.
RANK_REASON The cluster contains an academic paper detailing a new research framework for evolutionary program search.
Read on arXiv cs.NE (Neural & Evolutionary) →
- AgenticEvolve
- alphaXiv
- arXiv
- CatalyzeX Code Finder for Papers
- CORE Recommender
- DagsHub
- Gotit.pub
- Hugging Face
- No-analysis
- OpenEvolve
- ScienceCast
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →