CarryOn:为长程智能体构建面向对象的上下文
原标题:Paper page - Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents
AI 摘要
论文提出 CarryOn,一种面向长程智能体的面向对象上下文运行时,以可复用的任务对象而非交互历史来组织上下文。它通过延续载体、补充执行上下文和近期交互的实时尾部构建上下文视图,并在历史折叠后基于绑定证据和环境检查增量修订载体记录。在 OSWorld 2.0 和 BrowseComp-Plus 上,性能分别提升 5.8% 和 7.0%,总 token 用量分别减少 51.9% 和 22.9%。
正文节选
Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents Abstract CarryOn organizes long-horizon agent context around reusable task objects rather than interaction histories, improving efficiency and performance on complex benchmarks. Long-horizon agents accumulate progress in persistent task state, yet repeatedly process interaction histories describing how that progress was achieved. This raises a natural question: can long-horizon context instead be organized around reusa