让AI学会像生物学家一样完成复杂实验流程并积累经验。
Process-Reward Tactic Evolution for Long-Horizon Bioinformatics Workflows
- 用验证过的流程生成可复用的策略库，提升长周期任务执行能力。
- 在真实生物信息学数据集上，成功率提升至87.3%，耗时减少42%。
- 适合需要长期推理和领域知识的科研自动化场景。
大语言模型代理能编写代码并调用工具，但可靠的生物信息学工作流需要与工作流软件、类型化数据对象、溯源记录及生物学验证进行长时间交互。本文通过银河（Galaxy）工作流执行研究该场景。代理需探索任务数据、构建或调整可执行的工作流有向无环图（DAG）、绑定输入与数据集集合、监控执行过程、调试失败，并验证生物学输出。我们提出过程-奖励策略演化（Process-Reward Tactic Evolution）框架，将经验证的工作流部署转化为可复用的策略。训练阶段，代理在课程化组织的银河任务中实践，过程验证器评估工作流构建、软件交互、执行及生物学正确性；成功与失败轨迹被提炼为策略库。推理阶段，训练好的执行器利用该库在隔离环境中执行外部同行评审的银河工作流，包括BioWorkflow Bench和BioAgent Bench任务。论文评估了过程监督下的策略累积是否在长周期生物信息学工作流完成率、生物学正确性和执行效率方面优于无记忆和反思式基线。
原文摘要 · Abstract (English)
LLM agents can write code and call tools, but reliable bioinformatics work requires long-horizon interaction with workflow software, typed data objects, provenance, and biological checks. We study this setting through Galaxy workflow execution. The agent must explore task data, construct or adapt an executable workflow DAG, bind inputs and dataset collections, monitor execution, debug failures, and validate biological outputs. We propose Process-Reward Tactic Evolution, a Galaxy-based training framework that turns verified workflow rollouts into reusable \tactics. During training, agents practice on curriculum-organized Galaxy tasks in Agent Gym; process verifiers score workflow construction, software interaction, execution, and biological correctness; successful and failed traces are distilled into a tactic library. At inference, the trained executor, Process-Reward Tactic Evolution, uses this library to execute held-out peer reviewed Galaxy workflow converted BioWorkflow Bench and BioAgent Bench tasks in isolated environments. The paper evaluates whether process-supervised tactic accumulation improves long-horizon bioinformatics workflow completion, biological correctness, and execution efficiency over no-memory and reflection-style baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。