通过人类对齐设计用于药物发现中智能体AI的鲁棒性大语言模型评估系统

中文摘要 AI 辅助

智能体大语言模型（LLM）系统正在重塑化学与药物发现领域的科研工作流程，但对其开放式、工具增强型输出的评估仍是核心瓶颈。BLEU、ROUGE等基于参考的指标无法捕捉语义正确性，而专家人工评估无法匹配这些系统所需的迭代速度。LLM作为评判者（LLM-as-a-Judge）范式已成为一种可扩展的替代方案，但现有的药物发现基准部署LLM评判者时未验证其与人类专家的对齐性。本研究针对阿斯利康部署的智能体药物发现助手ChatInvent，提出一种LLM作为评判者的评估框架，包含四项贡献：第一，定义了四项输出质量评估维度——完整性、相关性、结构清晰度与范围一致性，同时辅以确定性工具调用正确性检查；第二，通过与五名专家标注者开展的人类对齐研究验证该评判者，对比Gemini 3.1 Pro、Claude Opus 4.7、GPT-5与Llama 3.1 70B作为候选评判者；第三，使用人工标注示例的少样本演示优化表现最佳的评判者，将其与人类多数投票的对齐度从0.80提升至0.86；第四，将优化后的评判者应用于70个保留问题，揭示具体局限性并发现非正式措辞不会系统性降低输出质量，相反，让LLM在查询智能体重写原始问题是有益的。本框架为科学领域智能体系统的人类对齐评估提供了可复用模板。

英文摘要

Agentic large language model (LLM) systems are reshaping scientific workflows in chemistry and drug discovery, but evaluating their open-ended, tool-augmented outputs remains a fundamental bottleneck. The LLM-as-a-Judge paradigm has emerged as a scalable alternative, but existing drug discovery benchmarks deploy LLM judges without validating their alignment with human experts. In this work, we present an LLM-as-a-Judge evaluation framework for ChatInvent, an agentic drug discovery assistant deployed at AstraZeneca, with five contributions. First, we define four output-quality evaluation dimensions---Completeness, Relevancy, Structural Clarity, and Scope Adherence---alongside deterministic Tool Call Correctness checks. Second, we validate the judge through a human alignment study with five expert annotators, comparing Gemini 3.1 Pro, Claude Opus 4.7, GPT-5, and Llama 3.1 70B as candidate judges. Third, we optimize the best-performing judge using few-shot demonstrations of human-annotated examples, improving alignment with the human majority vote from 0.80 to 0.86. Fourth, applying the optimized judge to 70 held-out questions, we surface concrete limitations and find no strong evidence that informal phrasing degrades output quality; it may, however, still be helpful to have the LLM rewrite the original question before querying the agent. Finally, we extend the framework to 38 adversarial questions that are ambiguous, invalid, out-of-scope or ethically sensitive, and show that the agent's refusal behavior is guided by the stated intent of a request. Our framework provides a reusable template for human-aligned evaluation of agentic systems in scientific domains.

发表机构

• AstraZeneca（阿斯利康）
• Chalmers University of Technology（查尔姆斯理工大学）
• University of Gothenburg（哥德堡大学）
• Science for Life Laboratory (SciLifeLab)（生命科学实验室）

机构由 AI 辅助整理，请以论文原文为准。