通过原子式结构动态调节智能体协作，提升效率与稳定性。
ATOM: Instantiating Budget-Controllable Multi-Agent Collaboration via Nucleus-Electron Hierarchy
- 用核心-电子层级结构，固定骨干+动态激活智能体
- 任务难度预估控制资源消耗，最高提升30%令牌效率
- 适合需要灵活调度计算资源的复杂多智能体场景
基于大语言模型的多智能体系统依赖优化的协作拓扑以平衡性能与通信开销。然而，现有方法在稳定性与可扩展性之间存在固有权衡，且常无法将计算预算与查询难度对齐。我们提出 extsc{ATOM}，一种通过任务驱动强化学习生成可预算调控协作图的自适应框架。受原子结构启发， extsc{ATOM} 采用核心-电子层级：维护一个稳定、离线学习的协作主干（核心），在推理时根据查询条件动态激活智能体（电子）。关键在于，一种复杂度感知的预算策略通过估计查询难度，严格调控电子实例化，实现资源消耗与任务需求的对齐。在六个多样化基准上的实验表明， extsc{ATOM} 实现了最先进性能，同时相比强基线提升最高达30%的令牌效率。
原文摘要 · Abstract (English)
Large Language Model (LLM)-based multi-agent systems rely on optimized collaboration topologies to balance performance and communication costs. However, current methods struggle with the inherent stability-extensibility trade-off and often misalign computational budgets with query difficulty. We propose \textsc{ATOM}, an adaptive framework that generates budget-controllable collaboration graphs via a novel task-driven reinforcement learning paradigm. Inspired by atomic structures, \textsc{ATOM} employs a nucleus-electron hierarchy: it maintains a stable, offline-learned collaboration backbone (the nucleus) while dynamically activating query-conditioned agents (electrons) during inference. Crucially, a complexity-aware budgeting strategy aligns resource consumption with task demands by estimating query difficulty to strictly regulate electron instantiation. Extensive experiments across six diverse benchmarks demonstrate that \textsc{ATOM} achieves state-of-the-art performance while improving token efficiency by up to $30\%$ compared to strong baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。