深度搜索能力已经成为前沿大语言模型(LLM)智能体,不可或缺的能力，但其发展仍然由工业巨头主导。典型的行业配方涉及资源高度密集的管道，涵盖预训练,持续预训练(CPT),监督微调(SFT),和强化学习(RL)。在本报告, 中，我们表明，当提供信息丰富且高难度的轨迹, 时，简单的 SFT 方法对于训练前沿搜索代理可能会非常强大。通过引入三个简单的数据合成修改: 缩放知识图大小以实现更丰富的探索, 扩展工具集大小以获得更广泛的功能, 和严格的低步过滤, 我们建立了更强大的基线。仅在 10.6k 数据点, 上进行训练，我们的 OpenSeeker-v2 在 4 个基准测试中实现了最先进的性能 (30B 大小的代理采用 ReAct 范例): 46.0% on BrowseComp, 58.1% on BrowseComp-ZH, 34.6% on Humanity的 Last Exam, 和xbench, 上的 78.0% 甚至超过了使用重型 CPT+SFT+RL pipeline, 训练的 Tongyi DeepResearch，分别达到了 43.4%,、46.7%,、32.9%, 和 75.0%,。值得注意的是，, OpenSeeker-v2 代表了其模型规模和范例中第一个最先进的搜索代理，由纯学术团队仅使用 SFT 开发。我们很高兴能够开源 OpenSeeker-v2 模型权重，并分享我们简单而有效的发现，以使社区更容易理解前沿搜索代理研究。
Deep search capabilities have become an indispensable competency for frontier Large Language Model (LLM) agents, yet their development remains dominated by industrial giants. The typical industry recipe involves a highly resource-intensive pipeline spanning pre-training, continual pre-training (CPT), supervised fine-tuning (SFT), and reinforcement learning (RL). In this report, we show that when fueled with informative and high-difficulty trajectories, a simple SFT approach could be surprisingly powerful for training frontier search agents. By introducing three simple data synthesis modifications: scaling knowledge graph size for richer exploration, expanding the tool set size for broader functionality, and strict low-step filtering, we establish a stronger baseline. Trained on merely 10.6k data points, our OpenSeeker-v2 achieves state-of-the-art performance across 4 benchmarks (30B-sized agents with ReAct paradigm): 46.0% on BrowseComp, 58.1% on BrowseComp-ZH, 34.6% on Humanity的 Last Exam, and 78.0% on xbench, surpassing even Tongyi DeepResearch trained with heavy CPT+SFT+RL pipeline, which achieves 43.4%, 46.7%, 32.9%, and 75.0%, respectively. Notably, OpenSeeker-v2 represents the first state-of-the-art search agent within its model scale and paradigm to be developed by a purely academic team using only SFT. We are excited to open-source the OpenSeeker-v2 model weights and share our simple yet effective findings to make frontier search agent research more accessible to the community.
科目: 人工智能 (cs.AI); 计算和语言 (cs.CL)
Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)