发表机构
Carnegie Mellon University; Meta（卡内基梅隆大学; Meta）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究长期智能体任务上下文管理问题，提出ACM框架，受人类记忆启发让智能体自主管理上下文，还开发训练后管道，能提升模型在相关任务上的性能，有效管理上下文可降低压力、实现扩展探索并产生更一致方案。
AI 中文摘要
智能体任务本质上是长期且多轮的，通过与环境交互不断积累上下文。现有上下文压缩方法不可避免地会导致信息丢失，且由僵化的启发式规则触发，与智能体不断演变的推理重点不一致。我们提出了智能体上下文管理（ACM）框架，为智能体配备专门的上下文编辑工具以进行无损上下文管理。受人类短期和长期记忆交互启发，智能体自主决定何时压缩上下文，将丢弃的内容卸载到外部存储系统，并按需查询以供后续检索。在此框架基础上，我们进一步开发了一个训练后管道，构建高质量的上下文管理示范，并提高智能体在搜索和编码任务上的模型性能。进一步分析表明，有效的上下文管理可降低峰值令牌压力，实现扩展探索，并在独立试验中产生更一致的解决方案。代码、数据和模型检查点可在该https网址获取。
英文摘要
Agentic tasks are inherently long-horizon and multi-turn, constantly accumulating context through interactions with the environment. Existing context compression methods inevitably incur information loss and are triggered by rigid heuristic rules, leaving them misaligned with the agent's evolving reasoning focus. We propose Agentic Context Management (ACM), a framework that equips agents with purpose-built context editing tools for lossless context management. Inspired by the interaction between short-term and long-term human memory, the agent autonomously decides when to compress its context, offloads discarded content to an external memory system, and queries it on demand for later retrieval. Building on this framework, we further develop a post-training pipeline that constructs high-quality demonstrations of context management and improves model performance on both agentic search and coding tasks. Further analysis reveals that effective context management reduces peak token pressure, enables extended explorations, and yields more consistent solutions across independent trials. Code, data, and model checkpoints are available at https://github.com/lixiaochuan2020/agentic-context-management.
