研究課題から科学ワークフローへ：科学自動化のためのエージェント型 AI 活用

arXiv:2604.21910v1 Announce Type: new Abstract: 科学ワークフローシステムは実行の自動化（スケジューリング、故障耐性、リソース管理）を担っています jedoch、それ以前に存在する意味的な変換は担っていません。科学家们仍然將研究課題手動轉換為工作流規範，這是一個需要領域知識和基礎設施專長的任務。我們提出了一種代理架構，通過三層來彌合這個差距：LLM 將自然語言解釋為結構化意圖（語義層）；驗證後的生成器產生可重複的有向無環圖（確定性層）；領域專家編寫「技能」：編碼詞彙映射、參數約束和最優化策略的文檔（知識層）。這種分解將 LLM 的非確定性限制在意圖提取中：相同的意圖始終產生相同的工作流。我們實現並在 1000 Genomes 人口遺傳學工作流和運行於 Kubernetes 上的 Hyperflow WMS 上評估了該架構。在 150 個查詢的消融研究中，技能將完全匹配意圖準確率從 44% 提高到 83%；技能驅動的工作流延遲生成將數據傳輸降低了 92%；端到端管道在 Kubernetes 上完成查詢，LLM 超額低於 15 秒，每次查詢成本低於 0.001 美元。

arXiv:2604.21910v1 Announce Type: new Abstract: Scientific workflow systems automate execution -- scheduling, fault tolerance, resource management -- but not the semantic translation that precedes it. Scientists still manually convert research questions into workflow specifications, a task requiring both domain knowledge and infrastructure expertise. We propose an agentic architecture that closes this gap through three layers: an LLM interprets natural language into structured intents (semantic layer); validated generators produce reproducible workflow DAGs (deterministic layer); and domain experts author ``Skills'': markdown documents encoding vocabulary mappings, parameter constraints, and optimization strategies (knowledge layer). This decomposition confines LLM non-determinism to intent extraction: identical intents always yield identical workflows. We implement and evaluate the architecture on the 1000 Genomes population genetics workflow and Hyperflow WMS running on Kubernetes. In an ablation study on 150 queries, Skills raise full-match intent accuracy from 44% to 83%; skill-driven deferred workflow generation reduces data transfer by 92\%; and the end-to-end pipeline completes queries on Kubernetes with LLM overhead below 15 seconds and cost under $0.001 per query.