发表机构
Shanghai Jiao Tong University; Microsoft Research; Purdue University（上海交通大学; 微软研究院; 普渡大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
CRAFT是一个模板感知的智能体框架，通过反思验证和受限局部修复提升电子表格填写准确性，在FormFillBench上显著优于基线。
AI 中文摘要
电子表格填写要求智能体整合外部证据、将值精确地定位到单元格，并保留不规则模板的结构。早期编辑中的错误可能覆盖标签或错位字段，从而影响后续决策。我们提出了CRAFT，一个模板感知的智能体框架，将反思性验证与受限的局部修复相结合。CRAFT不将反思视为重新生成工作簿的自由形式请求，而是将检测到的错误定位到电子表格区域，在必要时恢复损坏的模板状态，并在后续编辑前重新定位可写的槽位。一个矩形感知的槽位定位器（RASG）提出可写单元格，而标签-槽位提示和受保护区域约束后续编辑。我们引入了FormFillBench，包含327个表单，分为仅指令和多文件两个轨道。与最强的基线相比，CRAFT在这些轨道上的配对准确率分别提高了8.51和23.38个百分点。组件移除实验支持管道内的结构裁决和槽位重新定位，该框架在使用第二个骨干网络评估的方法中保持了其相对优势。代码和基准FormFillBench可在该https URL获取。
英文摘要
Spreadsheet form filling requires agents to consolidate external evidence, ground values to precise cells, and preserve irregular template structure. Errors in early edits can overwrite labels or misalign fields, undermining later decisions. We propose CRAFT, a template-aware agent framework that connects reflective validation to constrained local repair. Instead of treating reflection as a free-form request to regenerate the workbook, CRAFT grounds detected errors to spreadsheet regions, restores corrupted template state when necessary, and re-grounds plausible writable slots before subsequent edits. A Rectangle-Aware Slot Grounder (RASG) proposes writable cells, while label-slot hints and protected regions constrain subsequent edits. We introduce FormFillBench, with 327 forms across Instruction-Only and Multi-File tracks. Compared with the strongest baselines, CRAFT improves pair accuracy by 8.51 and 23.38 percentage points on these tracks, respectively. Component-removal experiments support structural adjudication and slot re-grounding within the pipeline, and the framework retains its relative advantage among the methods evaluated with a second backbone. The code and benchmark FormFillBench are available at https://github.com/Glllllly/CRAFT.
