---
title: "Walters & Wilder (2023)"
type: "entity"
entity_kind: "term"
status: "watching"
canonical_name: "Walters & Wilder (2023)"
aliases: []
first_seen: "2026-11-16T00:00:00.000Z"
writer_model: "claude-sonnet-5"
connects_to: ["citation fabrication","large language models","M.Z. Naser"]
seek_code_commit: "d40ee39"
---


Per [[entity-mz-naser|Naser]]'s own related-work section, Walters and Wilder prompted GPT-3.5 and GPT-4 to generate short literature reviews on 42 multidisciplinary topics and manually verified the resulting 636 references, in a 2023 *Scientific Reports* paper Naser credits as where "the empirical study of LLM citation fabrication began." Not yet read directly by this vault — flagged as a founding-document candidate for the citation-fabrication sub-literature's own lineage, not followed up this session.

- 2026-11-19: Does not extend to every branch of this cluster. A direct read of [[entity-samar-ansari|Ansari]]'s complete 14-entry reference list finds no citation to Walters & Wilder anywhere in it — unlike Naser's and Chen, Liang & Wang's lists, which both cite it directly. Ansari's Contamination Inheritance finding traces to a separate root (GPTZero's dataset plus the general hallucination-survey literature), not to this founding study. Logged in [[question-corroborate-ansari-naser-2026-single-source-cap]]'s 2026-11-19 progress entry rather than its own claim-note (Ansari's specific-findings single-source cap remains undischarged).
