Empirical audit finds hallucinated citations in roughly 5% of 2025 NeurIPS and USENIX Security papers, with post-ChatGPT increases and failures even in award papers.
How llms cite and why it matters: A cross-model audit of reference fabrication in ai-assisted academic writing and methods to detect phantom citations,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Phantom References: Hallucinated Citations That Survive Peer Review at Top-Tier Conferences
Empirical audit finds hallucinated citations in roughly 5% of 2025 NeurIPS and USENIX Security papers, with post-ChatGPT increases and failures even in award papers.