arXiv cs.AI· Junning Zhu, Peng He, Haohan Yuan, Jianpeng Liang·· 6 天前AI 评分24
GenCoord：私有信息下具身 LLM 智能体的可执行承诺协议
GenCoord: Executable Commitments for Embodied LLM Agents under Private Information
AI 导读
GenCoord 是一种去中心化协议，把分工本身当作消息：提案将发送方的技能路径与伙伴的后续工作配对为 Short DSL 表达，兼容路径形成承诺，并将履约检查与原始任务成功分开评估。
正文 · 原文
Abstract:Embodied language-model agents must coordinate interdependent actions while each holds private task information. In dialogue-based coordination, models generate messages and interpret replies to decide what each agent should do, spending inference before the team acts. We propose GenCoord, a decentralized protocol that makes the division of work itself the message. A proposal pairs the sender's skill path with its partner's continuation in Short DSL, a compact language for expressing joint work. Compatible paths form commitments linking the sender's delivery to the partner's work after verified receipt. We check fulfillment separately from original-task success. Complete requests also let a deterministic receiver proceed without another model call. In Minecraft, trained local 0.8B agents match the task success of controlled free-form communication while reducing median commitment time by 68.2%. The same-field full-key JSON control approaches free-form output length and delay, indicating that the amount generated matters even when messages are structured. With the same prompted DeepSeek model, both protocols complete nearly all tasks, but dialogue generates 2.8 times as many output tokens. Multistep commitments improve held-out-template success from 91.3% to 98.1% with fewer decisions. These results point to a design principle for embodied cooperation: make the division of work the message, and spend inference on choosing and revising the work.
| Comments: | Junning Zhu and Peng He contributed equally. Corresponding author: Peng He. Paper source and compact evidence: this https URL |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2608.22055 [cs.AI] |
| (or arXiv:2608.22055v3 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2608.22055 arXiv-issued DOI via DataCite |
Submission history
From: Junning Zhu [view email]
[v1]
Sat, 22 Aug 2026 17:41:25 UTC (1,851 KB)
[v2]
Tue, 25 Aug 2026 07:29:18 UTC (1,853 KB)
[v3]
Fri, 9 Oct 2026 15:20:03 UTC (1,698 KB)
来源：arXiv cs.AI · arxiv.org