“It is quite difficult to know for sure when content is AI generated. However, it turns out that AI really loves some specific names. If you find Elena Vasquez in the wild, she might be a real person. But if you find the story about Elena Vasquez and Marcus Chen together, we can be pretty confident that Claude made it up. The amazing couple have appeared a…”
The Ghost Couple: Correlated LLM Name Priors and Their Haunting of the Web and Academic Publishing
These names do not exist. Elena Vasquez and Marcus Chen have appeared as volcano experts, astronauts, thriller protagonists, podcast hosts, and academic co-authors across hundreds of independently produced AI-generated documents, never having lived. We show that large language models do not merely default to high-probability individual names when generating fictional experts: they produce correlated character ensembles, pairs and trios whose co-occurrence rates far exceed chance and are consistent across independent generations. These priors are model-family-specific (Claude: Elena Vasquez + Marcus Chen + Amara Okafor; Gemini: Aris Thorne + Lena Petrova; GPT: Elara Voss with no fixed partner), version-specific, and actively suppressed at model release boundaries, leaving dateable behavioral fingerprints in the content they produced. We document a downstream consequence at scale. On Zenodo, a CERN-operated repository that mints real DataCite DOIs, we identify 1,655 ghost-authored records claiming nonexistent journals with fabricated publication dates: server-side DataCite timestamps prove deliberate backdating, and 991 records were registered in a single month; these carry real DOIs registered in DataCite, making them harvestable by any scholarly aggregator that ingests DOI metadata. Ghost names additionally appear on ResearchGate forming synthetic research groups with collaborators drawn from multiple model families; publication dates on these records provide a reliable temporal proxy for model deployment windows.
候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。
HN 书评
没有评论直接链到它;下面是它被提交的讨论串里回复最多的评论(不计入统计)- LLMs use recurring ghost authors and personalities ↗Gaishan · 2026-06-12 · 2 分
- AI language models have favorite names, and we mapped them ↗mbrzozowski · 2026-06-15 · 1 分
- Correlated LLM Name Priors and Their Haunting of the Web and Academic Publishing ↗wise_blood · 2026-06-16 · 1 分
- Correlated LLM Name Priors and Their Haunting of the Web and Academic Publishing ↗sebg · 2026-06-18 · 1 分
- Ghost Couple:Correlated LLM Name Priors Haunting the Web and Academic Publishing ↗bookofjoe · 2026-09-01 · 1 分