RESEARCH INDEX
Practical research on GEO and AI search
This index separates public evidence from its limits and connects our papers, benchmark design, and field notes.
VeReach papers
Three papers on the structure that earns AI citations, evaluating GEO across whole conversations, and how far AI-based evaluation can be trusted. Each page covers the abstract, method, results, practical implications and limitations.
- GEO-SFE: structure drives AI citations
Structural Feature Engineering for Generative Engine Optimization: How Content Structure Shapes Citation Behavior
Junwei Yu, Mufeng Yang, Yepeng Ding, Hiroyuki Sato · March 31, 2026
Optimizing structure alone, without changing claims, lifted citation rate by +17.3% across six generative engines. Five principles for headings, paragraphs, lists, emphasis and internal links.
- Conversational Capture: evaluate GEO across the whole conversation
Conversational Capture: A Trajectory-Level Framework for Evaluating Generative Engine Optimization in Multi-turn Human-Agent Interaction
Junwei Yu, Jieyu Zhou, Mufeng Yang, Yepeng Ding, Hiroyuki Sato · September 30, 2026
Formalizes conversational capture, where an early-cited source keeps getting cited, and shows that single-turn evaluation can misrank GEO methods. HAI ’26 paper.
- JuryFlow: put LLM-judge disagreement to work
JuryFlow: Disagreement-Guided Human-in-the-Loop Multi-Agent Evaluation
Mufeng Yang, Junwei Yu, Yepeng Ding · September 30, 2026
Uses disagreement among LLM judges as an uncertainty signal for targeted re-evaluation, lifting LLMBar accuracy from 0.718 to 0.784 over majority vote. HAI ’26 paper.