Hallucination
A condition in which model-generated content can appear plausible yet is incorrect, unsupported, or inconsistent when assessed against relevant evidence, supplied context, or external facts.
Example
A model invents a paper title and author, then describes supposed findings as if the paper were real.
Distinctions and cautions
Hallucination does not refer to every type of AI error. Some literature separates inconsistencies with external facts from failures to remain faithful to the provided source or context. Factuality, Faithfulness, and Groundedness are related but not fully synonymous.
Open questions and cautions
There is no single definition of the scope of “Hallucination,” or single causal classification of how it arises, established across the research field.
Terminology note
“Confabulation” and “Fabrication” are sometimes used as related terms, but this Catalog does not currently treat them as exact synonyms for Hallucination.
Evidence
View evidence →
Survey of Hallucination in Natural Language Generation
- Title
- Survey of Hallucination in Natural Language Generation
- Author / organization
- Ji et al.
- Year
- 2023
- Source type / publication status
- ACM Computing Surveys major peer-reviewed survey; published
- Supported concepts / claim context
- Hallucination terminology/taxonomy; factuality/faithfulness boundary.
- Scope and limitations
- A major survey whose boundaries are deliberately broad to cover usage across tasks.
- Access / version note
- Published survey.
- Recorded official URL / DOI / identifier
- DOI:
10.1145/3571730(DOI resolver)
On Faithfulness and Factuality in Abstractive Summarization
- Title
- On Faithfulness and Factuality in Abstractive Summarization
- Author / organization
- Maynez et al.
- Year
- 2020
- Source type / publication status
- ACL peer-reviewed research; published
- Supported concepts / claim context
- Faithfulness, Factuality, Hallucination boundary.
- Scope and limitations
- Focused on summarization, but provides foundational evidence for the source-versus-world distinction.
- Access / version note
- Published ACL.
- Recorded official URL / DOI / identifier
- DOI:
10.18653/v1/2020.acl-main.173(DOI resolver)
Evaluating large language models for accuracy incentivizes hallucinations
- Title
- Evaluating large language models for accuracy incentivizes hallucinations
- Author / organization
- Kalai et al.
- Year
- 2026
- Source type / publication status
- Nature peer-reviewed research; published
- Supported concepts / claim context
- Accuracy, Abstention, Hallucination; evaluation incentives.
- Scope and limitations
- Formal model and selected frontier-model case study. Utility depends on the use case.
- Access / version note
- Published Nature article.
- Recorded official URL / DOI / identifier
- DOI:
10.1038/s41586-026-10549-w(DOI resolver)
Detecting hallucinations in large language models using semantic entropy
- Title
- Detecting hallucinations in large language models using semantic entropy
- Author / organization
- Farquhar et al.
- Year
- 2024
- Source type / publication status
- Nature peer-reviewed research; published
- Supported concepts / claim context
- Uncertainty Estimation, Abstention, Hallucination.
- Scope and limitations
- Requires sampling and clustering by semantic equivalence; cost and applicability depend on the setting.
- Access / version note
- Published Nature article.
- Recorded official URL / DOI / identifier
- DOI:
10.1038/s41586-024-07421-0(DOI resolver)
Chain-of-Verification Reduces Hallucination in Large Language Models
- Title
- Chain-of-Verification Reduces Hallucination in Large Language Models
- Author / organization
- Dhuliawala et al.
- Year
- 2024
- Source type / publication status
- Findings ACL peer-reviewed research; published
- Supported concepts / claim context
- Verification; mitigation of hallucination.
- Scope and limitations
- Self-/model-based verification does not guarantee external ground truth.
- Access / version note
- Published ACL Findings.
- Recorded official URL / DOI / identifier
- DOI:
10.18653/v1/2024.findings-acl.212(DOI resolver)
Verify with Caution: The Pitfalls of Relying on Imperfect Factuality Metrics
- Title
- Verify with Caution: The Pitfalls of Relying on Imperfect Factuality Metrics
- Author / organization
- Godbole & Jia
- Year
- 2025
- Source type / publication status
- Findings ACL peer-reviewed research; published
- Supported concepts / claim context
- Limits of verification; disagreement among factuality metrics.
- Scope and limitations
- Depends on the factuality metrics and datasets studied.
- Access / version note
- Published ACL Findings.
- Recorded official URL / DOI / identifier
- DOI:
10.18653/v1/2025.findings-acl.1175(DOI resolver)
Astute RAG: Overcoming Imperfect Retrieval Augmentation and Knowledge Conflicts for Large Language Models
- Title
- Astute RAG: Overcoming Imperfect Retrieval Augmentation and Knowledge Conflicts for Large Language Models
- Author / organization
- Wang et al.
- Year
- 2025
- Source type / publication status
- ACL peer-reviewed research; published
- Supported concepts / claim context
- Retrieval Failure, RAG, Knowledge Conflict.
- Scope and limitations
- Controlled RAG analyses; no universal extrapolation to all retrievers or domains.
- Access / version note
- Published ACL.
- Recorded official URL / DOI / identifier
- DOI:
10.18653/v1/2025.acl-long.1476(DOI resolver)