FUSE : 多模态搜索与推荐中子代理证据的故障感知使用
FUSE : Failure-aware Usage of Subagent Evidence for MultiModal Search and Recommendation
AI总结：
FUSE通过上下文压缩策略提升多模态搜索与推荐性能，实现93.3%的意图准确率和99.4%的召回率。
摘要：
多模态创意助手将用户目标分解并路由到子代理以进行布局、样式、检索和生成。检索质量至关重要，但故障可能在多个阶段出现：理解用户意图、选择内容类型、寻找候选（召回）或排序结果。同时，发送和处理图像成本高昂，使朴素的多模态方法不切实际。我们提出了FUSE：多模态搜索与推荐中子代理证据的故障感知使用。FUSE将大部分原始图像提示替换为紧凑的Grounded Design Representation（GDR）：一个由规划团队提供的canvas元素（图像、文本、形状、图标、视频、logo）、结构、样式、显著颜色和用户选择组成的意识选择JSON。FUSE实现了七种上下文预算策略：全面的基础提示、上下文压缩、推理链、小样本优化、检索增强上下文、两阶段处理和零样本极简主义。最后，一个管道归因层通过将子代理信号转换为简单的检查来监控系统性能：意图对齐、内容类型/路由合理性、召回健康（例如零命中和顶级匹配强度）以及排序位移分析。我们评估了七种上下文预算变体，通过788个来自多样化用户和设计模板的评估查询（参考图3）。我们的系统评估表明，上下文压缩在所有管道阶段均取得最佳性能，93.3%的意图准确率、86.8%的路由成功率（含回退）、99.4%的召回率和88.5%的NDCG@5。这种方法表明，战略上下文总结优于全面和极简上下文化策略。
英文摘要：
Multimodal creative assistants decompose user goals and route tasks to subagents for layout, styling, retrieval, and generation. Retrieval quality is pivotal, yet failures can arise at several stages: understanding user intent, choosing content types, finding candidates (recall), or ranking results. Meanwhile, sending and processing images is costly, making naive multimodal approaches impractical. We present FUSE: Failure-aware Usage of Subagent Evidence for MultiModal Search and Recommendation. FUSE replaces most raw-image prompting with a compact Grounded Design Representation (GDR): a selection aware JSON of canvas elements (image, text, shape, icon, video, logo), structure, styles, salient colors, and user selection provided by the Planner team. FUSE implements seven context budgeting strategies: comprehensive baseline prompting, context compression, chain-of-thought reasoning, mini-shot optimization, retrieval-augmented context, two-stage processing, and zero-shot minimalism. Finally, a pipeline attribution layer monitors system performance by converting subagent signals into simple checks: intent alignment, content-type/routing sanity, recall health (e.g., zero-hit and top-match strength), and ranking displacement analysis. We evaluate the seven context budgeting variants across 788 evaluation queries from diverse users and design templates (refer Figure 3). Our systematic evaluation reveals that Context Compression achieves optimal performance across all pipeline stages, with 93.3% intent accuracy, 86.8% routing success(with fallbacks), 99.4% recall, and 88.5% NDCG@5. This approach demonstrates that strategic context summarization outperforms both comprehensive and minimal contextualization strategies.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
