An Explication and Classroom Field Study of the Virtual Human Interaction Lab’s Expert (VHIL-E) LLM
Bibliographic Data
| ID | 17820985 |
|---|---|
| Authors | Jeremy N Bailenson (0000-0003-2813-3297, Stanford University, corresponding author), Jonathan You (Stanford University), D M Markowitz (0000-0002-7159-7014, Michigan State University), Gustav Petersen (Stanford University), Rabindra Ratan (0000-0001-7611-8046, Michigan State University), Monique Santoso (0000-0002-7513-3196, Stanford University), Portia Wang (0000-0003-1704-5718, Stanford University) |
| Year | 2026 |
| Volume | 29 |
| Issue | 3 |
| Pages | 177-183 |
| Publication date | 2026-03-01 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Cyberpsychology Behavior and Social Networking (JOURNAL) |
| Journal identifiers | ISSN: 2152-2715 • E-ISSN: 2152-2723 |
| Publisher | Mary Ann Liebert, Inc (PUBLISHER • US) |
| DOI | 10.1177/21522715261423752 |
| PMID | 41766249 |
| OpenAlex | W7133227328 |
| Language | EN |
| References cited | 17 |
89) used VHIL-E to query and understand the course materials and logged hallucinations, which were "egregiously wrong answers." Students then chose the single worst wrong answer over 8 weeks. They then compared the Base/RAG Hybrid-which prioritized the RAG but allowed ChatGPT to consult its general intelligence-to the RAG Constrained, which limited substantive information to the embedded index. Allowing VHIL-E access to GPT produced more than twice as many hallucinations as constraining it to the index. We discuss implications for scholars who build and use RAG-based LLM applications
Curriculum · Explication · Implementation · Outreach · Virtual actor · AI in Service Interactions · Artificial Intelligence in Healthcare and Education · Intelligent Tutoring Systems and Adaptive Learning
| Citation velocity | historical |
|---|---|
| Highly cited | No |