TreasureX Researchcs.AIAutomating Scientific Figure Editing with AI That Learns from Paper Revisions
Yasheng Sun, Zezi Zeng, Yifan Yang, Chong Luo, Wenyi Wang, Ziwei Liu, Jürgen Schmidhuber · July 16, 2026
Editing figures in research papers is tedious but essential. A new system, SciDiagramEdit, learns to automate this by analyzing before-and-after figure pairs from arXiv paper revisions. It uses an AI agent that improves its editing skills over time through trial and error, following natural language instructions like "relabel this axis" or "rearrange these panels." The key innovation is that it works on vector graphics, allowing precise edits to individual elements (text, arrows, plots) while keeping the figure's overall structure intact. This could save researchers hours of manual work, especially during multiple revision rounds. The system's ability to learn from real revision histories makes it practical and scalable. For investors, this points to growing demand for AI tools in academic publishing and research workflow automation. As scientific output grows, automating routine tasks like figure editing becomes a valuable niche. The paper shows that even complex, domain-specific editing can be taught to AI using natural revision data.
Read more: https://arxiv.org/abs/2607.15272
Educational content only, generated with AI assistance from public research abstracts. Not investment advice. Thank you to arXiv for use of its open access interoperability.