ViviDoc: Generating Interactive Documents through Human-Agent Collaboration

2026-03-29

AI Summary

This work addresses the challenge of interactive document authoring, which demands both domain expertise and front-end development capabilities—tasks that existing large language model approaches struggle to control with sufficient precision. The paper presents the first systematic solution through a multi-agent collaborative framework encompassing planning, styling, execution, and evaluation agents. Key innovations include the SRTC interaction protocol (State-Rendering-Transition-Constraint), a content-aware style palette, and a multi-granularity human-AI co-editing mechanism. The authors introduce ViviBench, the first benchmark comprising 101 diverse topics, and evaluate their approach via a four-dimensional automated metric suite alongside user studies. Experimental results demonstrate significant improvements over baseline methods in both content richness and interaction quality, with validation from 12 users confirming high usability, strong controllability, and overall satisfaction.

Abstract

Interactive documents help readers engage with complex ideas through dynamic visualization, interactive animations, and exploratory interfaces. However, creating such documents remains costly, as it requires both domain expertise and web development skills. Recent Large Language Model (LLM)-based agents can automate content creation, but directly applying them to interactive document generation often produces outputs that are difficult to control. To address this, we present ViviDoc, to the best of our knowledge the first work to systematically address interactive document generation. ViviDoc introduces a multi-agent pipeline (Planner, Styler, Executor, Evaluator). To make the generation process controllable, we provide three levels of human control: (1) the Document Specification (DocSpec) with SRTC Interaction Specifications (State, Render, Transition, Constraint) for structured planning, (2) a content-aware Style Palette for customizing writing and interaction styles, and (3) chat-based editing for iterative refinement. We also construct ViviBench, a benchmark of 101 topics derived from real-world interactive documents across 11 domains, along with a taxonomy of 8 interaction types and a 4-dimensional automated evaluation framework validated against human ratings (Pearson r > 0.84). Experiments show that ViviDoc achieves the highest content richness and interaction quality in both automated and human evaluation. A 12-person user study confirms that the system is easy to use, provides effective control over the generation process, and produces documents that satisfy users.

Problem

• interactive documents
• human-agent collaboration
• controllable generation
• document authoring
• Large Language Models

Innovation

• interactive document generation
• multi-agent system
• human-agent collaboration
• controllable generation
• ViviBench