将反应知识图谱接地于大语言模型以实现合成检索
Grounding Large Language Models in Reaction Knowledge Graphs for Synthesis Retrieval
AI总结：
本研究通过将反应路径检索转化为图查询生成问题，利用对齐示例的单样本提示提升大语言模型在合成规划中的检索准确性。
摘要：
大语言模型（LLMs）可以辅助化学合成规划，但标准提示方法常常产生幻觉或过时的建议。我们通过将反应路径检索视为文本到图查询（Text2Cypher，自然语言到图查询）生成问题，研究LLM与反应知识图谱的交互。我们定义了单步和多步检索任务。我们通过使用静态、随机和嵌入式示例选择，比较零样本提示与单样本变体，并评估基于检查表的验证/校正循环。为了评估我们的框架，我们考虑查询有效性和检索准确性。我们发现，使用对齐示例的单样本提示表现最佳。我们的检查表式自校正循环主要在零样本设置中提高可执行性，并在已有良好示例时提供有限的额外检索增益。我们提供了一个可重复的Text2Cypher评估设置，以促进进一步的KG接地LLM在合成规划中的研究。代码可在https://github.com/Intelligent-molecular-systems/KG-LLM-Synthesis-Retrieval获取。
英文摘要：
Large Language Models (LLMs) can aid synthesis planning in chemistry, but standard prompting methods often yield hallucinated or outdated suggestions. We study LLM interactions with a reaction knowledge graph by casting reaction path retrieval as a Text2Cypher (natural language to graph query) generation problem, and define single- and multi-step retrieval tasks. We compare zero-shot prompting to one-shot variants using static, random, and embedding-based exemplar selection, and assess a checklist-driven validator/corrector loop. To evaluate our framework, we consider query validity and retrieval accuracy. We find that one-shot prompting with aligned exemplars consistently performs best. Our checklist-style self-correction loop mainly improves executability in zero-shot settings and offers limited additional retrieval gains once a good exemplar is present. We provide a reproducible Text2Cypher evaluation setup to facilitate further work on KG-grounded LLMs for synthesis planning. Code is available at https://github.com/Intelligent-molecular-systems/KG-LLM-Synthesis-Retrieval.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
