arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.11964cs.AI
收藏

Gaia2:在动态和异步环境中评估大语言模型代理的基准测试

Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments

Meta SuperIntelligence Labs(Meta超级智能实验室)
Romain Froger, Pierre Andrews, Matteo Bettini, Amar Budhiraja, Ricardo Silveira Cabral, Virginie Do, Emilien Garreau, Jean-Baptiste Gaya, Hugo Laurençon, Maxime… 展开作者

Romain Froger, Pierre Andrews, Matteo Bettini, Amar Budhiraja, Ricardo Silveira Cabral, Virginie Do, Emilien Garreau, Jean-Baptiste Gaya, Hugo Laurençon, Maxime Lecanu, Kunal Malkan, Dheeraj Mekala, Pierre Ménard, Gerard Moreno-Torres Bertran, Ulyana Piterbarg, Mikhail Plekhanov, Mathieu Rita, Andrey Rusakov, Vladislav Vorotilov, Mengjue Wang, Ian Yu, Amine Benhalloum, Grégoire Mialon, Thomas Scialom

AI总结:

Gaia2通过动态和异步环境评估大语言模型代理,揭示了推理、效率和鲁棒性之间的权衡,为代理系统的发展提供灵活的基础设施。

摘要:

我们介绍了Gaia2,一个用于评估大语言模型代理在现实中的异步环境中的基准测试。不同于以往静态或同步的评估,Gaia2引入了环境独立于代理动作演变的场景,要求代理在时间约束下操作,适应噪声和动态事件,解决歧义,并与其他代理协作。每个场景都配有一个写入操作验证器,使Gaia2能够实现细粒度、动作级别的评估,并使其直接可用于从可验证奖励中学习。我们对当前最先进的专有和开源模型的评估显示,没有模型在所有能力上占据主导:GPT-5(高)达到了最强的总体分数42% pass@1,但在时间敏感任务上失败,Claude-4 Sonnet在准确性和速度上牺牲了成本,Kimi-K2在开源模型中领先,达到21% pass@1。这些结果突显了推理、效率、鲁棒性之间的根本权衡,并暴露了缩小“sim2real”差距的挑战。Gaia2基于消费者环境,使用开源的Agents Research Environments平台构建,并设计为易于扩展。通过将Gaia2与基础的ARE框架一起发布,我们旨在为社区提供一个灵活的基础设施,用于开发、基准测试和训练下一代实用代理系统。

英文摘要:

We introduce Gaia2, a benchmark for evaluating large language model agents in realistic, asynchronous environments. Unlike prior static or synchronous evaluations, Gaia2 introduces scenarios where environments evolve independently of agent actions, requiring agents to operate under temporal constraints, adapt to noisy and dynamic events, resolve ambiguity, and collaborate with other agents. Each scenario is paired with a write-action verifier, enabling fine-grained, action-level evaluation and making Gaia2 directly usable for reinforcement learning from verifiable rewards. Our evaluation of state-of-the-art proprietary and open-source models shows that no model dominates across capabilities: GPT-5 (high) reaches the strongest overall score of 42% pass@1 but fails on time-sensitive tasks, Claude-4 Sonnet trades accuracy and speed for cost, Kimi-K2 leads among open-source models with 21% pass@1. These results highlight fundamental trade-offs between reasoning, efficiency, robustness, and expose challenges in closing the "sim2real" gap. Gaia2 is built on a consumer environment with the open-source Agents Research Environments platform and designed to be easy to extend. By releasing Gaia2 alongside the foundational ARE framework, we aim to provide the community with a flexible infrastructure for developing, benchmarking, and training the next generation of practical agent systems.

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。

↑