收录解读
这篇论文把 deep search agent 数据构造中的 shortcut 问题形式化。
它不仅增加结构复杂度，而是控制证据覆盖、线索选择性、暴露常数和先验绑定等真实捷径风险。
它值得收录，因为 shortcut-resistant search task synthesis 是训练和评估深搜索 agent 的基础设施。
局限在于当前证据主要来自预印本实验与作者自建评测，后续需要独立复现和更大范围部署验证。
原始摘要与中文对照
中文对照翻译
FORT-Searcher：合成抗捷径搜索任务以训练深度搜索智能体。训练深度搜索智能体需要可验证的问题，这些问题的答案在通过搜索获取到足够证据之前是不可用的。现有的合成方法通常通过丰富图结构来增加表观难度，但仅凭结构复杂性并不能保证实际的搜索难度：预期的搜索过程可能通过更廉价的识别路径而崩溃。我们用一个捷径感知难度框架来形式化这一差距，并识别出四种可操作的捷径风险：证据共覆盖、单线索选择性、暴露常数和先验知识绑定。为了诊断它们的实际影响，我们使用轨迹特征，包括解决成本、答案命中时间和先验捷径率。在此框架的指导下，我们引入了 F O RT，一个抗捷径训练数据合成框架。F O RT 通过在实体选择、证据图构建、问题制定和对抗性细化中控制捷径风险来构建抗捷径训练数据。实验表明，与现有开源深度搜索数据集相比，FORT 能够诱导更长的预答案搜索和更少的捷径模式。利用由此产生的轨迹，我们仅使用监督微调（SFT）训练了 F O RT-Searcher，它在具有挑战性的深度搜索基准上，在同等规模的开源搜索智能体中取得了最佳的整体性能。相关资源将在 https://github.com/RUCAIBox/FORT-Searcher 提供。
原始摘要
Training deep search agents requires verifiable questions whose answers remain unavailable until sufficient evidence has been acquired through search. Existing synthesis methods often increase apparent difficulty by enriching graph structures, but structural complexity alone does not guarantee realized search difficulty: the intended search process can collapse through a cheaper identifying route. We formalize this gap with a shortcut-aware difficulty framework and identify four actionable shortcut risks: evidence co-coverage, single-clue selectivity, exposed constants, and prior-knowledge binding. To diagnose their realized effects, we use trajectory signatures including solving cost, answer hit time, and prior-shortcut rate. Guided by this framework, we introduce F O RT, a Framework of Shortcut-Resistant Training-Data Synthesis. F O RT constructs shortcut-resistant training data by controlling shortcut risks across entity selection, evidence graph construction, question formulation, and adversarial refinement. Experiments show that FORT induces longer pre-answer search and fewer shortcut patterns than existing open-source deep search datasets. Using the resulting trajectories, we train F O RT-Searcher with supervised finetuning (SFT) only, and it achieves the best overall performance among comparable-size open-source search agents on challenging deep search benchmarks. Relevant resources will be made available at https://github.com/RUCAIBox/FORT-Searcher. FORTSearcher