AI学会自己教自己搜索：不用老师，自己练出超强搜索力
现在的AI搜索代理大多靠“老师”教——要么模仿专家操作，要么靠稀疏的奖励信号摸索，但遇到多步搜索任务就容易卡住。这篇论文反其道而行：先造了一个完全可控的“搜索实验室”（DeepSearch-World），里面所有搜索和阅读结果都是确定可重复的，然后让AI自己反复尝试、筛选成功路径、再自我微调，循环迭代。结果，一个9B参数的模型，在没有更强模型指导的情况下，在BrowseComp（复杂网页搜索）上达到31.2%，在GAIA（通用AI助手）上61.5%，在HotpotQA（多跳问答）上93.4%，逼近甚至超越许多依赖外部蒸馏的模型。它不是你明天就能用的工具，但揭示了一个趋势：只要环境足够“诚实”（可验证、可重复），AI可以像人类一样从试错中自学，不再依赖昂贵的专家数据。
📄 原文摘要(英文)
Training tool-use agents to improve from their own experience remains challenging, as supervised fine-tuning relies on fixed teacher-distilled trajectories, while sparse-reward reinforcement learning provides weak supervision for long-horizon interactions. We present DeepSearch-Evolve, a self-distillation framework for web agents built on DeepSearch-World, a deterministic and verifiable environment with reproducible search and page-reading tools. DeepSearch-World contains 420K multi-hop QA tasks constructed from entity-level random walks and supports key agentic cognitive behaviors useful for self-evolving, including progress verification, grounded reflection, and failure recovery. DeepSearch-Evolve iteratively performs trajectory generation, filtering, data mixing, and fine-tuning to train stronger agents. Without distillation from more capable models, DeepSearch-World-9B achieves competitive performance compared with open-source agents, reaching 31.2% on BrowseComp, 61.5% on GAIA, and 93.4% on HotpotQA, showing that verifiable environments enable scalable self-evolution for long-horizon web agents. We will release the environment, 420K training pool, validation set, model, and code to facilitate future research on self-improving deep search agents.