AI 中文总结
该研究提出智能体AI框架，结合LLM推理与RAG实现遗传编程父代选择配置自动化，经符号回归测试，5 mini--AR配置表现优于锦标赛选择，为进化系统自动化设计提供了新路径。
AI 中文摘要
我们研究智能体人工智能是否能通过引入智能体框架来自动完成遗传编程系统设计的部分流程，该框架利用大语言模型（LLM）的推理和检索增强生成技术识别并实现父代选择算法。以符号回归为测试平台，我们首先针对四种LLM类型开展 ablation 研究，评估智能体推理和检索对生成算法类别、有效性、实现相似度及下游性能的影响。结果显示，这些组件会显著影响生成算法的类型，但其下游性能很大程度上取决于底层LLM。最强配置为包含5 mini（5 mini--AR）的完整智能体设置，该配置持续生成已确立的ε-词典序选择实现，同时保持有竞争力的下游性能。随后我们将该配置与锦标赛选择、半动态MAD ε-词典序选择的固定实现进行基准测试，在六个符号回归问题上，5 mini--AR的表现与ε-词典序选择相近，且总体优于锦标赛选择。这些发现证明了智能体人工智能将领域知识转化为可执行组件的潜力，为进化系统的自动化配置与设计迈出了一步。
英文摘要
We investigate whether agentic artificial intelligence can automate parts of the process of designing genetic programming systems by introducing an agentic framework that identifies and implements parent selection algorithms using large language model (LLM) reasoning and retrieval-augmented generation. Using symbolic regression as a test bed, we first conduct an ablation study across four LLM types to evaluate the effects of agentic reasoning and retrieval on generated algorithm categories, validity, implementation similarity, and downstream performance. Results show that these components substantially influence the types of algorithms generated, but their downstream performance largely depends on the underlying LLM. The strongest configuration, the full agentic setup with 5 mini (5 mini--AR), consistently generated established $ε$-lexicase implementations while maintaining competitive downstream performance. We then benchmark this configuration against fixed implementations of tournament selection and semi-dynamic MAD $ε$-lexicase. Across six symbolic regression problems, 5 mini--AR performed similarly to $ε$-lexicase while generally outperforming tournament selection. These findings demonstrate the potential of agentic AI to translate domain knowledge into generating executable components, providing a step toward automated configuration and design of evolutionary systems.
CommentsUpdated Benchmarking figure with correct statistical test results
