面向动态推理的修订感知独立智能体图

中文摘要 AI 辅助

传统推理协议呈现的是固定的、预先选择的任务，因此无法测试智能体是否传播相关更新、保留未受影响的工作，或重建历史任务绑定。为此，我们研究动态任务路由，其中事件流修订任务绑定，系统必须在求解前选择每个查询时间点有效的文档版本。为研究此问题，我们将六个广泛使用的基准（MMLU、MMLU-Pro、MedMCQA、MATH、GPQA 和 HumanEval）改造为包含 373,428 个按时间分类查询的 31,119 个动态片段。该设置揭示了一个核心权衡：每次事件后重新计算会浪费工作，而无保护的复用则返回过时结论。我们引入了修订感知独立智能体图（RIAG），这是一种有界的多智能体策略，将确定性时间解析与任务推理分离。RIAG 按不可变文档标识缓存解决方案，每个新任务以两次未暴露的尝试开始，并有条件地调用审计和修复，每个文档版本最多使用四次调用。在此集合上，同构 RIAG 在 0.62 次调用/查询下实现了 54.24% 的联合路由和回答准确率，而最强比较方法在 18.00 次调用/查询下仅为 32.22%；异构 RIAG 在 0.63 次调用/查询下达到 49.78%。

英文摘要

Conventional reasoning protocols present a fixed, preselected task, so they cannot test whether an agent propagates relevant updates, preserves unaffected work, or reconstructs a historical task binding. We therefore study \emph{dynamic task routing}, in which an event stream revises task bindings and a system must select the document version valid at each query time before solving it. To study this problem, we repurpose six widely used benchmarks: MMLU, MMLU-Pro, MedMCQA, MATH, GPQA, and HumanEval into 31{,}119 dynamic episodes comprising 373{,}428 temporally categorized queries. This setting exposes a central trade-off: recomputing after every event wastes work, whereas unguarded reuse returns stale conclusions. We introduce the Revision-Aware Independent Agent Graph (RIAG), a bounded multi-agent policy that separates deterministic temporal resolution from task reasoning. RIAG caches solutions by immutable document identity, starts each fresh task with two unexposed attempts, and conditionally invokes audit and repair, using at most four calls per document version. On this collection, homogeneous RIAG achieves 54.24\% joint routing-and-answer accuracy at 0.62 calls/query, compared with 32.22\% at 18.00 calls/query for the strongest comparison method; heterogeneous RIAG reaches 49.78\% at 0.63 calls/query.

发表机构

• Harvard AI and Robotics Lab（哈佛人工智能与机器人实验室）
• Harvard University（哈佛大学）

机构由 AI 辅助整理，请以论文原文为准。