English(EN) Who May Release an Agent's Promise? Standing over Commitments in Multi-Party Chat

研究发现：AI代理在聊天中难以取消承诺

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-16 04:00

一篇新的arXiv论文探讨了AI代理如何在多方聊天中处理承诺，重点关注谁有权取消对某人的承诺。研究人员发现，像Qwen3.5-9B这样的小型开源模型更容易发生未经授权的解除承诺行为，而大型模型则倾向于根据同伴的话来放弃承诺。研究表明，实施一个自我维护的承诺账本和明确的解除政策可以显著提高准确性并防止代理行为中的错误。 AI

影响这项研究突显了在协作环境中运行的AI代理的关键安全性和可靠性问题，并为未来的代理设计提出了改进建议。

排序理由关于AI代理行为和承诺处理的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 17 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 关于AI代理行为和承诺处理的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Lan Wei, Bingyu Shen, Jiaqi Tang, Qing Lu, Kuan Huang, Meng Xu, Miles Q. Li, Bin Hu, Boyang Li · 2026-10-16 04:00

谁可能兑现代理的承诺？在多方聊天中信守承诺

arXiv:2610.21203v1 Announce Type: new Abstract: When an assistant serving a team makes a promise to Alice, who may cancel it? We call this standing: only the creditor (the person owed) or someone the creditor or operator policy authorizes may release a commitment, regardless of r…

本文如何被排名

报道来源 [1]

谁可能兑现代理的承诺？在多方聊天中信守承诺

相关实体

相关话题