用低成本方法生成2万条可验证的搜索推理数据，提升小模型搜索能力
ORBIT: Scalable and Verifiable Data Generation for Search Agents on a Tight Budget
- 四阶段框架自动生成带验证的复杂查询数据，无需付费API
- 2万条数据覆盖15个领域，每条需4-5步推理且需全网验证
- 在小模型上表现超群，适合资源有限的研究者复现
搜索代理将语言模型与网络搜索结合，成为解答复杂用户问题的关键。然而，构建涉及多步检索与推理的深度研究训练数据仍面临人工标注成本高、前置条件繁琐等挑战。本文提出ORBIT，一个包含20,000条推理密集型查询的训练数据集，采用低成本框架生成，不依赖付费API服务。该模块化框架包含四个阶段：种子创建、问答对生成，以及自验证和外部验证两阶段验证机制。ORBIT覆盖15个领域，每条训练样本需4-5步推理，外部验证需基于完整网络内容。我们使用GRPO方法在ORBIT上训练了Qwen3-4B基座模型，并在维基百科问答任务上进行评估。大量实验结果表明，ORBIT-4B在参数量小于40亿的模型中表现优异，证明了合成数据集的有效性。相关框架、代码与数据集均已开源，可供公开获取。
原文摘要 · Abstract (English)
Search agents, which integrate language models (LMs) with web search, are becoming crucial for answering complex user queries. Constructing training datasets for deep research tasks, involving multi-step retrieval and reasoning, remains challenging due to expensive human annotation, or cumbersome prerequisites. In this work, we introduce ORBIT, a training dataset with 20K reasoning-intensive queries with short verifiable answers, generated using a frugal framework without relying on paid API services. The modular framework relies on four stages: seed creation, question-answer pair generation, and two stages of verification: self and external. ORBIT spans 15 domains and each training pair requires 4-5 reasoning steps, with external search verification required from the complete web. We train Qwen3-4B as the base model on ORBIT using GRPO and evaluate it on Wikipedia question answering tasks. Extensive experiment results demonstrate that ORBIT-4B achieves strong performance among sub-4B LLMs as search agents, proving the utility of synthetic datasets. Our framework, code and datasets are open-sourced and available publicly.
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。