GPTZero Hallucination Check Tool Identifies 50+ Citation Errors in ICLR 2026 Submissions
By
puttycat
Hot, fresh, and worth queueing round the block for.
Summary
GPTZero's Hallucination Check tool identified over 50 instances of hallucinations in papers under review for ICLR 2026 that were missed by 3-5 peer reviewers each. The article presents a table showing examples of flagged papers, including one titled 'TamperTok: Forensics-Driven Tokenized Autoregressive Framework for Image Tampering Localization' with an 8.0 rating, where GPTZero detected a hallucination in the citation section. The tool appears to be used for academic paper verification, specifically checking for fabricated or inaccurate citations and references in machine learning research submissions.
Key quotes
· 4 pulledGPTZero used our Hallucination Check tool to find 50+ hallucinations under review at ICLR, each of which were missed by 3-5 peer reviewers.
TamperTok: Forensics-Driven Tokenized Autoregressive Framework for Image Tampering Localization | OpenReview
Chong Zou, Zhipeng Wang, Ziyu Li, Nan Wu, Yuling Cai, Shan Shi, Jiawei Wei, Xia Sun, Jian Wang, and Yizhou Wang. Segment everything everywhere all at once. In Advances in Neural Information Processing Systems (NeurIPS),
Example of Verified Hallucination
You might also wanna read
Columbia professor studying AI nearly publishes paper with AI-hallucinated reference, highlighting growing academic integrity concern
A Columbia University associate professor, Maxim Topaz, who researches AI applications in healthcare, nearly published a scientific paper co
exacted.me·4d ago
Columbia researcher's own paper nearly published with AI-hallucinated reference, highlighting growing academic integrity concern
A Columbia University associate professor, Maxim Topaz, who researches AI applications in healthcare, nearly published a scientific paper co
fortune.com·4d ago
OpenAI says GPT-5.5 Instant reduces ChatGPT hallucinations by over 50% on high-stakes prompts
OpenAI claims its new GPT-5.5 Instant model, now the default for ChatGPT, hallucinates significantly less than the previous GPT-5.3 Instant
The Verge·26d ago
AI-generated research papers overwhelm academic peer review and citation systems
The article discusses a growing crisis in academic publishing where AI-generated research papers are flooding journals and citation database
The Verge·17d ago
New benchmark reveals AI models often cite wrong sources even when answers are correct
Researchers at Peking University have developed CiteVQA, a new benchmark that tests whether AI models can correctly cite source documents wh
the-decoder.com·4d ago