通过持续预训练构建强代理基础模型，提升开源代理的自主推理与工具使用能力。
Scaling Agents via Continual Pre-training
- 引入持续预训练机制，专为代理行为建模优化基础模型
- 在10个基准上达到顶尖性能，中文浏览任务达43.3%准确率
- 适合需要高自主性、多步骤推理的研究型智能体开发者
大语言模型已演变为具备自主工具调用与多步推理能力的代理系统，可解决复杂问题。然而，基于通用基础模型的后训练方法在代理任务中表现不佳，尤其在开源实现中。我们发现根本原因在于缺乏稳健的代理基础模型，导致后训练阶段需同时学习多样代理行为并对其对齐专家示范，引发优化冲突。为此，我们首次将代理持续预训练（Agentic CPT）引入深度研究代理训练流程，构建强大的代理基础模型。基于此，我们开发了名为AgentFounder的深度研究代理模型。在10个基准上评估AgentFounder-30B，表现达到业界领先水平，同时保持强大工具使用能力，其中在BrowseComp-en上达39.9%，BrowseComp-zh上达43.3%，HLE任务中Pass@1为31.5%。
原文摘要 · Abstract (English)
Large language models (LLMs) have evolved into agentic systems capable of autonomous tool use and multi-step reasoning for complex problem-solving. However, post-training approaches building upon general-purpose foundation models consistently underperform in agentic tasks, particularly in open-source implementations. We identify the root cause: the absence of robust agentic foundation models forces models during post-training to simultaneously learn diverse agentic behaviors while aligning them to expert demonstrations, thereby creating fundamental optimization tensions. To this end, we are the first to propose incorporating Agentic Continual Pre-training (Agentic CPT) into the deep research agents training pipeline to build powerful agentic foundational models. Based on this approach, we develop a deep research agent model named AgentFounder. We evaluate our AgentFounder-30B on 10 benchmarks and achieve state-of-the-art performance while retains strong tool-use ability, notably 39.9% on BrowseComp-en, 43.3% on BrowseComp-zh, and 31.5% Pass@1 on HLE.
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。