不要重启：面向有状态智能体交接的残差补全
Do Not Restart: Residual Completion for Stateful Agent Handoffs
查看机构详情
- Nanjing University（南京大学）
- Singapore Management University（新加坡管理大学）
机构由 AI 辅助整理，请以论文原文为准。
浏览论文内容
中文总结 AI 辅助
针对有状态智能体交接中需保留已接受选项和未完成义务的问题，提出承诺前沿残差补全（CFRC）方法，通过冻结残差契约并验证实时证据来保证正确性，在五个环境中以22.0%-34.6%的推理成本达到与全任务智能体相当的准确率。
中文摘要 AI 辅助
路由和级联通过跨模型转移控制来降低工具智能体的成本，但有状态的交接必须保留已接受的选项、已实现的效果和未完成的义务。我们将此形式化为受承诺约束的残差补全，并引入承诺前沿残差补全（CFRC）。CFRC 强制执行目标先于提案、完整提案先于授权、实时证据先于成功：它从已接受的进展中冻结残差契约，将后继延续闭合为证据关联图，并且仅在剩余部分被覆盖且实时收据履行义务时才允许执行。我们建立了契约相对的部分正确性，在完整契约构造下，该性质可推广到原始残差请求。在五个环境和两个同提供商模型对上，CFRC 实现了与强全任务智能体相当的宏平均准确率，而其推理成本仅为后者的22.0%-34.6%，额外的跨提供商结果展示了更广泛的迁移能力。
英文摘要
Routing and cascades reduce tool-agent cost by transferring control across models, but stateful handoffs must preserve accepted choices, realized effects, and unfinished obligations. We formulate this as commitment-constrained residual completion and introduce Commitment-Frontier Residual Completion (CFRC). CFRC enforces target-before-proposal, whole-proposal-before-authority, and live-evidence-before-success: it freezes a residual contract from accepted progress, closes the successor continuation into an evidence-linked graph, and admits execution only when the remainder is covered, with live receipts discharging obligations. We establish contract-relative partial correctness, which extends to the original residual request under complete contract construction. Across five environments and two same-provider model pairs, CFRC achieves comparable macro accuracy to strong full-task agents at 22.0% to 34.6% mean per-surface cost, with additional cross-provider results demonstrating broader transfer.
