AgentSwing：面向长视野网络智能体的自适应并行上下文管理路由技术

摘要

随着大语言模型（LLMs）逐渐演变为面向长期信息搜索的自主智能体，有限上下文容量的管理已成为关键瓶颈。现有上下文管理方法通常在整个任务轨迹中采用单一固定策略，此类静态设计在某些状态下可能表现良好，但无法适应长期搜索过程中累积上下文的有用性与可靠性动态演变的特性。为系统化这一挑战，我们提出了一个概率框架，通过搜索效率与终端精度这两个互补维度来刻画长期任务的成功机制。基于此视角，我们提出AgentSwing——一种状态感知的自适应并行上下文管理路由框架。在每个触发点，AgentSwing并行扩展多个上下文管理分支，并通过前瞻式路由选择最具潜力的延续路径。在多样化基准测试和智能体骨干网络上的实验表明，AgentSwing始终优于强力的静态上下文管理方法，通常能以最多减少3倍的交互轮次达到或超越其性能，同时提升长期网络智能体的终极性能上限。除实证优势外，该概率框架为分析和设计面向长期智能体的上下文管理策略提供了理论视角。

English

As large language models (LLMs) evolve into autonomous agents for long-horizon information-seeking, managing finite context capacity has become a critical bottleneck. Existing context management methods typically commit to a single fixed strategy throughout the entire trajectory. Such static designs may work well in some states, but they cannot adapt as the usefulness and reliability of the accumulated context evolve during long-horizon search. To formalize this challenge, we introduce a probabilistic framework that characterizes long-horizon success through two complementary dimensions: search efficiency and terminal precision. Building on this perspective, we propose AgentSwing, a state-aware adaptive parallel context management routing framework. At each trigger point, AgentSwing expands multiple context-managed branches in parallel and uses lookahead routing to select the most promising continuation. Experiments across diverse benchmarks and agent backbones show that AgentSwing consistently outperforms strong static context management methods, often matching or exceeding their performance with up to 3times fewer interaction turns while also improving the ultimate performance ceiling of long-horizon web agents. Beyond the empirical gains, the proposed probabilistic framework provides a principled lens for analyzing and designing future context management strategies for long-horizon agents.

AgentSwing：面向长视野网络智能体的自适应并行上下文管理路由技术

AgentSwing: Adaptive Parallel Context Management Routing for Long-Horizon Web Agents

摘要

Support