Масштабируя горизонт, а не параметры: достижение производительности на уровне триллиона параметров с помощью 35-миллиардного агента
Scaling the Horizon, Not the Parameters: Reaching Trillion-Parameter Performance with a 35B Agent
June 29, 2026
Авторы: Lei Bai, Zongsheng Cao, Yang Chen, Zhiyao Cui, Shangheng Du, Yue Fan, Shiyang Feng, Zijie Guo, Haonan He, Liang He, Xiaohan He, Shuyue Hu, Yusong Hu, Songtao Huang, Yichen Jiang, Hao Li, Xin Li, Dahua Lin, Weihao Lin, Fenghua Ling, Dongrui Liu, Zhuo Liu, Runmin Ma, Chunjiang Mu, Haoyang Peng, Tianshuo Peng, Jinxin Shi, Luohe Shi, Boyuan Sun, Zelin Tan, Shengji Tang, Qianyi Wang, Yiming Wu, Yi Xie, Xiangchao Yan, Jingqi Ye, Peng Ye, Fangchen Yu, Jiakang Yuan, Bihao Zhan, Bo Zhang, Chen Zhang, Shufei Zhang, Shuaiyu Zhang, Wenlong Zhang, Yiqun Zhang, Junpeng Zhao, Zhijie Zhong, Bowen Zhou, Yuhao Zhou
cs.AI
Аннотация
我们提出Agents-A1——一个35B参数的混合专家智能体模型，通过扩展智能体视野达到万亿参数级别的性能。我们从两个维度研究智能体视野扩展：长视野轨迹的扩展与异构智能体能力的扩展。为实现这一目标，我们构建了长视野知识-行动基础设施，将外部知识、行动、观测与验证器结果相连接，生成平均长度为45K tokens的智能体轨迹。在此基础上，我们采用三阶段方案训练Agents-A1：首先开展全领域有监督微调，使基础模型适应广泛的智能体行为；其次训练领域级教师模型以捕捉各领域的专业专长；最后提出一种多教师领域路由在线蒸馏方法，结合显著词汇对齐技术，提升跨领域知识迁移效率，将六个异构领域统一整合至一个可部署的学生模型。Agents-A1在长视野智能体基准测试中展现出强大而广泛的性能。与万亿参数模型（如Kimi-K2.6和DeepSeek-V4-pro）相比，Agents-A1在SEAL-0（56.4）、IFBench（80.6）、HiPhO（46.4）、FrontierScience-Olympiad（79.0）和MolBench-Bind（56.8）上取得领先结果，同时在SciCode（44.3）、HLE（47.6）和BrowseComp（75.5）上保持高度竞争力。我们希望这项工作为社区提供一条切实可行的路径，通过35B参数智能体在长视野任务上达到或匹敌万亿参数模型的性能。
English
We introduce Agents-A1, a 35B Mixture-of-Experts Agentic Model that reaches trillion-parameter-level performance by scaling the agent horizon. We investigate agent-horizon scaling from two perspectives: scaling long-horizon trajectories and scaling heterogeneous agent abilities. To support this goal, we build a long-horizon knowledge-action infrastructure that connects external knowledge, actions, observations, and verifier outcomes, producing agentic trajectories with an average length of 45K tokens. Based on this, we train Agents-A1 with a three-stage recipe. First, we perform full-domain supervised fine-tuning to align the base model with broad agentic behaviors. Second, we train domain-level teacher models to capture specialized expertise in each domain. Third, we propose a multi-teacher domain-routed on-policy distillation with salient vocabulary alignment to improve knowledge transfer efficiency across different domains, unifying six heterogeneous domains into one deployable student model. Agents-A1 achieves strong and broad performance for long-horizon agent benchmarks. Compared with 1T-parameter model such as Kimi-K2.6 and DeepSeek-V4-pro, Agents-A1 achieves leading results on SEAL-0 (56.4), IFBench (80.6), HiPhO (46.4), FrontierScience-Olympiad (79.0), and MolBench-Bind (56.8), and remains highly competitive on SciCode (44.3), HLE (47.6) and BrowseComp (75.5). We hope this work provides the community with a practical path for scaling the horizon using a 35B agent that can reach or match the performance of 1T models on long-horizon tasks.