论文

FLUID框架:让AI任务指导更灵活,学会适时沉默

FLUID: Flexible, Evidence-Grounded Situated Task Guidance

精选理由

这篇论文解决了一个容易被忽略的问题:AI实时指导你做事时,什么时候该说话、什么时候该闭嘴。它用真实的人人指导数据替代合成对话来评估系统,还给出了四项指标的综合基准结果。做语音助手、AR眼镜指导或具身交互的团队可以重点看看评估框架。

研究团队推出FLUID框架,用于建模和评估多模态任务指导中的灵活交互。该框架基于WTaG和HoloAssist两个自然人与人指导数据集,将AI干预锚定在支撑证据、必要性和时机弹性三个维度上。

正文

传统基准把合成的对话当作固定答案,只比对内容是否一致,忽略了同一时刻多种回应都合理、有时保持沉默才是最优选择这一现实。FLUID用统一的表征方式同时完成两件事:评估指导系统的灵活性,以及构建一个实时任务指导Agent。

作者报告其Agent在回应正确性、干预精准度、指导时机和响应延迟四项指标的综合平衡上优于所有被测系统,但在避免语音重叠方面仍有提升空间。这项工作把情境化任务指导从僵硬的参照匹配推向更自然的自适应交互。

原文 · arXiv cs.AI

FLUID: Flexible, Evidence-Grounded Situated Task Guidance

Real-time task guidance is inherently flexible: when an agent guides a user through a task, multiple responses and timings can be appropriate, and sometimes the best action is to remain silent. Existing benchmarks often miss this flexibility by treating synthesized dialogues as fixed targets for what to say and when to say it. We introduce FLUID (Flexible, Evidence-Grounded Situated Task Guidance), a unified framework for modeling, evaluating, and enabling flexible interaction in multimodal task guidance. Using naturally occurring human-human guidance from WTaG and HoloAssist, FLUID grounds interventions in their supporting evidence, necessity, and timing flexibility, and uses this representation both to evaluate flexible guidance and to build a real-time agent for task guidance. Our agent achieves the strongest overall balance of response correctness, intervention precision, timely guidance, and response latency among evaluated systems, while still leaving room for improvement in avoiding speech overlap. Together, FLUID moves situated task guidance beyond rigid reference matching toward more natural and adaptive human-AI interaction. Project page: https://fluid-guidance.github.io.

产业坐标与信源核验

本条情报归属 AITOP「论文」核心栏目,重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:FLUID、任务指导、多模态、智能体,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。

信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Yunqi Zhao, Run Peng, Ze Dong, Ryan Lu, Yinpei Dai, Joyce Chai),经 AITOP 多维评分管线结构化提炼。