动态组合异构模型与工具，提升跨领域复杂推理能力
Atlas: Orchestrating Heterogeneous Models and Tools for Multi-Domain Complex Reasoning
- 基于经验先验的无训练聚类路由，实现领域自适应匹配
- 强化学习驱动多步路由，在分布外任务上性能提升13.1%
- 适合需要跨域推理和多模态工具协同的AI系统开发者
将大语言模型（LLMs）与外部工具结合显著拓展了AI代理的能力。然而，随着模型和工具种类增多，选择最优模型-工具组合成为高维优化难题。现有方法常依赖单一模型或固定调用逻辑，未能利用异构组合间的性能差异。本文提出ATLAS（自适应工具-模型对齐与协同调用），一种双路径动态工具使用框架，用于跨域复杂推理。该框架包含：(1) 无需训练的基于聚类的路由，利用领域特定的经验先验；(2) 基于强化学习的多步路由，探索分布外泛化路径。在15个基准上的实验证明，该方法超越闭源模型GPT-4o，对分布内任务提升10.1%，分布外任务提升13.1%。此外，通过协调专用多模态工具，在视觉推理任务中也取得显著提升。
原文摘要 · Abstract (English)
The integration of large language models (LLMs) with external tools has significantly expanded the capabilities of AI agents. However, as the diversity of both LLMs and tools increases, selecting the optimal model-tool combination becomes a high-dimensional optimization challenge. Existing approaches often rely on a single model or fixed tool-calling logic, failing to exploit the performance variations across heterogeneous model-tool pairs. In this paper, we present ATLAS (Adaptive Tool-LLM Alignment and Synergistic Invocation), a dual-path framework for dynamic tool usage in cross-domain complex reasoning. ATLAS operates via a dual-path approach: (1) \textbf{training-free cluster-based routing} that exploits empirical priors for domain-specific alignment, and (2) \textbf{RL-based multi-step routing} that explores autonomous trajectories for out-of-distribution generalization. Extensive experiments across 15 benchmarks demonstrate that our method outperforms closed-source models like GPT-4o, surpassing existing routing methods on both in-distribution (+10.1%) and out-of-distribution (+13.1%) tasks. Furthermore, our framework shows significant gains in visual reasoning by orchestrating specialized multi-modal tools.
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。