DAGent: Evaluate-then-Grow Planning for Deep Research Agents

• 类型：arxiv
• 标识：2609.39154
• 链接：https://arxiv.org/abs/2609.39154
• 主分类：agent
• 形态：method
• TLDR：Deep research tasks require agents to navigate large knowledge spaces, synthesize evidence across many sources, and adapt their plans as findings emerge. Directed acyclic graph (DAG)-based multi-agent systems suit this setting because they support parallel execution and isolate each sub-task within a focused dependency context. Yet existing DAG-based agents instantiate a task-level plan before execution and repair the graph only after failures or missing evidence are observed. This Plan-then-Patch strategy is brittle for deep research: the system commits most strongly when its evidence is weak
• 副分类：evaluation
• 待LLM分类：否
• 标题中文：DAGent：面向深度研究 Agent 的 Evaluate-then-Grow 规划
• TLDR中文：深度研究任务要求 Agent 在大规模知识空间中导航、跨多源证据进行综合合成，并根据发现调整计划。基于有向无环图（DAG）的多 Agent 系统天然适合该场景，因其支持并行执行并将每个子任务隔离在聚焦的依赖上下文中。然而现有基于 DAG 的 Agent 在执行前实例化任务级计划，仅在失败或证据缺失后才修补图结构。这种 Plan-then-Patch 策略对深度研究而言十分脆弱：当证据最薄弱时系统反而最为确定……
• 来源文件：
• /inbox/tom/_candidates/2026-10-02-agent-rag-longcontext-candidates.json