arXiv 2505.202862025-05-26cs.AI

Alita：通用智能体，以最小预定义和最大自我进化实现可扩展的智能体推理

Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolution

Princeton University（普林斯顿大学）；Tsinghua University（清华大学）；Shanghai Jiao Tong University（上海交通大学）；University of Michigan（密歇根大学）；Tianqiao and Chrissy Chen Institute（天桥和克里斯·陈研究所）；The Chinese University of Hong Kong（香港中文大学）

Jiahao Qiu, Xuan Qi, Tongcheng Zhang, Xinzhe Juan, Jiacheng Guo, Yifu Lu, Yimin Wang, Zixin Yao, Qihan Ren, Xun Jiang, Xing Zhou, Dongrui Liu, Ling Yang, Yue Wu… 展开作者

Jiahao Qiu, Xuan Qi, Tongcheng Zhang, Xinzhe Juan, Jiacheng Guo, Yifu Lu, Yimin Wang, Zixin Yao, Qihan Ren, Xun Jiang, Xing Zhou, Dongrui Liu, Ling Yang, Yue Wu, Kaixuan Huang, Shilong Liu, Hongru Wang, Mengdi Wang

备注：: 9 pages, 3 figures

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

Alita是一个通用智能体，通过最小预定义和最大自我进化实现可扩展的智能体推理，在GAIA、Mathvista和PathVQA上取得领先性能。

摘要：

近年来，大型语言模型（LLMs）的进展使得智能体能够自主执行复杂的、开放式任务。然而，许多现有框架严重依赖手动预定义的工具和工作流，这阻碍了它们跨领域的适应性、可扩展性和泛化能力。在这项工作中，我们引入了Alita——一个遵循“简单是终极 sophistication”原则设计的通用智能体，通过最小预定义和最大自我进化实现可扩展的智能体推理。对于最小预定义，Alita仅配备一个用于直接解决问题的组件，使其比以往严重依赖手工制作、精心设计的工具和工作流的方法更简单、更整洁。这种简洁的设计增强了其泛化到挑战性问题的潜力，而不受工具限制。对于最大自我进化，我们通过提供一套通用组件来激发Alita的创造力，使其能够通过从开源生成与任务相关的模型上下文协议（MCPs）来自主构建、完善和重用外部能力，这有助于可扩展的智能体推理。值得注意的是，Alita在GAIA基准验证数据集上达到了75.15%的pass@1和87.27%的pass@3准确率，在通用智能体中排名靠前；在Mathvista和PathVQA上分别达到74.00%和52.00%的pass@1，优于许多复杂度高得多的智能体系统。更多细节将在https://github.com/CharlesQ9/Alita更新。

英文摘要：

Recent advances in large language models (LLMs) have enabled agents to autonomously perform complex, open-ended tasks. However, many existing frameworks depend heavily on manually predefined tools and workflows, which hinder their adaptability, scalability, and generalization across domains. In this work, we introduce Alita--a generalist agent designed with the principle of "Simplicity is the ultimate sophistication," enabling scalable agentic reasoning through minimal predefinition and maximal self-evolution. For minimal predefinition, Alita is equipped with only one component for direct problem-solving, making it much simpler and neater than previous approaches that relied heavily on hand-crafted, elaborate tools and workflows. This clean design enhances its potential to generalize to challenging questions, without being limited by tools. For Maximal self-evolution, we enable the creativity of Alita by providing a suite of general-purpose components to autonomously construct, refine, and reuse external capabilities by generating task-related model context protocols (MCPs) from open source, which contributes to scalable agentic reasoning. Notably, Alita achieves 75.15% pass@1 and 87.27% pass@3 accuracy, which is top-ranking among general-purpose agents, on the GAIA benchmark validation dataset, 74.00% and 52.00% pass@1, respectively, on Mathvista and PathVQA, outperforming many agent systems with far greater complexity. More details will be updated at $\href{https://github.com/CharlesQ9/Alita}{https://github.com/CharlesQ9/Alita}$.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。