研究库 论文知识库

内容库 / 论文

arXiv:2610.05709 · Agent 智能体

Nexus: An Execution Fabric for AI Agents Across Cloud, Edge, and Devices

Nexus: An Execution Fabric for AI Agents Across Cloud, Edge, and Devices

类型：arxiv
标识：2610.05709
链接：http://arxiv.org/abs/2610.05709v1
主分类：agent
形态：method
TLDR：Language-model agents are evolving into long-running services that interact with models, tools, computers, mobile devices, and distributed environments. Existing agent frameworks simplify reasoning and tool invocation. However, cloud-centric designs face three limitations: centralized execution increases failure impact, scaling pressure, and compute cost; extending agents across computers, mobile devices, and edge environments requires a unified execution abstraction with permission control; and long-running executions require consistent lifecycle management across failures, recovery, results,
待LLM分类：否
来源文件：
/inbox/tom/_candidates/2026-10-06-agent-rag-longcontext-candidates.json

相关 · 按语义匹配

笔记 Interconnects (Nathan Lambert) · RSS 摘要

笔记 CSDN 高价值 AI 研究检索 · 2026-07-30

仓库 devloperdevesh/EnterpriseRAG-AI

笔记 Tom 文献雷达 Addendum · AI Agent · RAG · 长上下文 · 2026-06-28

笔记研究知识库草稿 · Jay · 2026-06-14 下午

仓库 Azure/aks-mcp

arXiv

2610.05709

TLDR

Language-model agents are evolving into long-running services that interact with models, tools, computers, mobile devices, and distributed environments. Existing agent frameworks simplify reasoning and tool invocation. However, cloud-centric designs face three limitations: centralized execution increases failure impact, scaling pressure, and compute cost; extending agents across computers, mobile devices, and edge environments requires a unified execution abstraction with permission control; and long-running executions require consistent lifecycle management across failures, recovery, results,

arXiv ↗ 中文全文