# AURA： 面向隐式需求的定向探测方法

- 来源：HuggingFace Daily Papers（社区热门论文）
- 发布时间：2026-06-04 08:00
- AIHOT 分数：59
- AIHOT 链接：https://aihot.virxact.com/items/cmq13ksuq0c4ysltr8yqdeknr
- 原文链接：https://arxiv.org/abs/2606.05557

## AI 摘要

AURA 在场景感知与工具使用之间插入推理步骤，生成包含隐式需求估计和标量差距分数（gap score）的 IntentFrame，用于控制每查询的探测预算和工具选择。在 100 查询四场景隐式意图基准上，AURA 相比 ReAct 风格探测将隐式需求覆盖率提升 0.07（p < 10⁻⁶），其中三个场景统计显著，且在第二个骨干模型上复现；消融实验将提升归因于差距校准而非答案记忆。在事实查找任务中，控制器以 82% 更少的探测次数和隐私敏感片段零违规换取原始准确率。代码、模拟器和基准已开源。

## 正文

A situated query like "where is Lin Wei?" often encodes more than its literal content: the user may also want to know whether Lin Wei is free, in a good mood, or worth interrupting now. Standard tool-use agents answer the literal question and stop. AURA inserts an inference step between scene perception and tool use that produces an IntentFrame: a structured estimate of the implicit need with a scalar gap score that controls per-query probe budget and tool selection. On a 100-query four-scene implicit-intent benchmark, AURA improves implicit-need coverage over ReAct-style probing (Delta = +0.07, p < 10^-6); three of four scenes are individually significant, the gain reproduces on a second backbone, and a prompt ablation attributes the lift to gap calibration rather than answer memorisation. On factual lookup the controller trades raw accuracy for 82% fewer probes and zero forbidden-tool violations on a privacy-sensitive slice; scope conditions are detailed in Limitations. Code, simulator, and benchmark are released at https://github.com/innovation64/AURA.