Policy Maps:引导大语言模型行为无界空间的工具
Policy Maps: Tools for Guiding the Unbounded Space of LLM Behaviors
AI总结:
针对大语言模型行为空间庞大、AI策略覆盖难的问题,提出受地图制作启发的policy maps方法及交互式工具Policy Projector,支持策略可视化编写,经专家验证可有效制定问题行为管控策略。
摘要:
AI策略为AI模型的可接受行为设定边界,但在大语言模型(LLM)场景下这颇具挑战:如何确保覆盖范围遍及庞大的行为空间?我们提出policy maps(策略地图)——一种受实体地图制作实践启发的AI策略设计方法。策略地图不以实现全面覆盖为目标,而是通过精心设计选择要捕捉的内容和要抽象忽略的内容,来辅助有效导航。借助Policy Projector(策略投影仪)——一款用于设计LLM策略地图的交互式工具,AI从业者可以查看模型输入-输出对的全貌,定义自定义区域(例如“暴力”),并通过可作用于LLM输出的if-then策略规则在这些区域中导航(例如,若输出包含“暴力”和“血腥细节”,则重写以移除“血腥细节”)。Policy Projector依托LLM分类与引导技术,支持交互式策略编写,并提供反映AI从业者工作成果的地图可视化界面。在一项由12名AI安全专家参与的评估中,我们的系统可帮助策略设计者围绕错误性别假设、即时人身安全威胁处理等问题模型行为制定策略。
英文摘要:
AI policy sets boundaries on acceptable behavior for AI models, but this is challenging in the context of large language models (LLMs): how do you ensure coverage over a vast behavior space? We introduce policy maps, an approach to AI policy design inspired by the practice of physical mapmaking. Instead of aiming for full coverage, policy maps aid effective navigation through intentional design choices about which aspects to capture and which to abstract away. With Policy Projector, an interactive tool for designing LLM policy maps, an AI practitioner can survey the landscape of model input-output pairs, define custom regions (e.g., "violence"), and navigate these regions with if-then policy rules that can act on LLM outputs (e.g., if output contains "violence" and "graphic details," then rewrite without "graphic details"). Policy Projector supports interactive policy authoring using LLM classification and steering and a map visualization reflecting the AI practitioner's work. In an evaluation with 12 AI safety experts, our system helps policy designers craft policies around problematic model behaviors such as incorrect gender assumptions and handling of immediate physical safety threats.
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。