Metraipolis › archive › dc_7f43eb63f989

The Labeling Problem in Hallucination Detection Benchmarks: An Empirical Evaluation — arXiv:2610.08026

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.08026
License (as archived)
CC0-1.0
Archived
2026-10-07T11:23:18.328Z
SHA-256 of the archived text
5de77cdb2e14f11e9de606de1ca8f8d5708cbcaa5265833597ede8a87cfe5f74
Stable URL of this page
https://metraipolis.com/documento/dc_7f43eb63f989

How to cite

The Labeling Problem in Hallucination Detection Benchmarks: An Empirical Evaluation — arXiv:2610.08026 — arXiv (en), https://arxiv.org/abs/2610.08026 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7f43eb63f989 (Metraipolis archive dc_7f43eb63f989, sha256 5de77cdb2e14f11e9de606de1ca8f8d5708cbcaa5265833597ede8a87cfe5f74)
BibTeX
@misc{dc7f43eb63f989,
  title = {The Labeling Problem in Hallucination Detection Benchmarks: An Empirical Evaluation — arXiv:2610.08026},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.08026},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7f43eb63f989; sha256 5de77cdb2e14f11e9de606de1ca8f8d5708cbcaa5265833597ede8a87cfe5f74},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.08026. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

In recent years, several methods for detecting when large language models (LLMs) hallucinate have been developed. These methods are often benchmarked with open-domain question answering (QA) datasets containing questions and corresponding short reference answers. First, an LLM is used to generate answers to questions within the QA dataset. Then, some automated labeling strategy is used to label these answers as hallucinated or not by comparing them with the reference answers in the dataset. This evaluation setting creates a methodological ambiguity between two criteria: reference faithfulness (whether the answer is fully supported by the reference) and factual correctness (whether the answer is free from contradictions and factually false specific claims). In practice, automated labelers may apply the former criterion even when the intended target is the latter. We study this potential criterion mismatch using 900 human-labeled question-answer pairs spanning three commonly used QA datasets and three generator models, with labels targeting answer-level factual correctness. We evaluate lexical similarity metrics, a reference-entailment NLI baseline, and seven LLM judges under controlled prompt variants as automated labelers. Our experiments reveal substantial disagreement both among automated labeling strategies and between these labels and human annotations. Many strategies also exhibit strong directional error biases, and for most judge-generator pairs, replacing a faithfulness-oriented prompt with a factual-correctness prompt improves agreement with human annotations and reduces false-positive dominance, indicating that automated hallucination labels depend strongly on how the target criterion is specified. Label-source choice should therefore be considered a fundamental part of benchmark design and made explicit, validated, and matched with the benchmark goal.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.08026 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T11:15:39.444Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).