搜索可教之外：智能体视觉生成中知识边界的演变

摘要

视觉生成器擅长渲染，却会自信地编造自身未知的内容。用户需求无穷无尽、持续演变且深度呈长尾分布：新角色、热门实体、截止日期后的事件等层出不穷。这种世界知识瓶颈在结构上是根本性的：生成器基于固定语料库训练，而视觉世界却是开放式的。我们构建了 SearchGen-20K 和 SearchGen-Bench 数据集，包含涵盖十二类故障模式、二十二个领域的20,839条提示词，并配套提供预执行的 SearchGen-Corpus-1M 多模态语料库，以支持离线、可复现的研究。在 SearchGen-Bench 上，前沿开源生成器满分100分仅得21至28分，较现有基准出现40分的断崖式下跌。自然的解决方案是引入搜索工具，实现智能体式视觉生成。然而我们发现，简单搜索并不可靠：它会无差别检索，向生成器原本能处理的提示词中注入噪声。我们追踪到根本原因在于生成器特有的、动态变化的知识边界——即生成器通过训练能够内化的知识与必须保留在外部上下文中的知识之间的分界线。虽然这一边界难以预先界定，但我们证明它可以通过"先教学后搜索"的协同训练框架来发现。即便是该协同训练方案的最小实现也能产生单调改进，为视觉生成领域实现递归式自我改进、满足世界知识驱动的需求奠定基础。我们发布完整数据集、协同训练语料库和搜索语料库，作为面向工具增强、基于世界知识的视觉生成的可复现测试平台。

English

Visual generators excel at rendering, but they confidently fabricate what they do not know. User requests are unbounded, evolving, and deeply long-tailed: new characters, trending entities, post-cutoff events, and more. This world-knowledge bottleneck is structural: generators are trained on fixed corpora, but the visual world is open-ended. We construct SearchGen-20K and SearchGen-Bench, with 20,839 prompts spanning twelve failure categories and twenty-two domains, paired with a pre-executed multimodal SearchGen-Corpus-1M to support offline, reproducible research. On SearchGen-Bench, frontier open generators score only 21 to 28 out of 100, a 40-point collapse invisible to existing benchmarks. The natural remedy is to employ search tools, enabling agentic visual generation. However, we find that naive search fails: it retrieves indiscriminately, injecting noise into prompts the generator already handles. We trace the root cause to a generator-specific, evolving knowledge boundary: the divide between what a generator can internalize through training and what must remain in external context. Although this boundary is hard to specify in advance, we show that it is discoverable through a teach-then-search co-training framework. Even a minimal version of this co-training recipe produces monotonic improvement, laying the foundation for recursive self-improvement in visual generation that can meet world-knowledge-grounded requests. We release the full dataset, co-training corpus, and search corpus as a replayable harness for tool-augmented, world-knowledge-grounded visual generation.