Metraipolis › archive › dc_6087442e48cc

Fact over Fiction: Detection of Pathological Hallucinations in Sinhala-to-English Neural Machine Translation — arXiv:2610.11389

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.11389
License (as archived)
CC0-1.0
Archived
2026-10-09T06:06:17.516Z
SHA-256 of the archived text
a515d2450559892c68c22652ef6d563477d6c1dff5ff2f1a717b6568f950bfed
Stable URL of this page
https://metraipolis.com/documento/dc_6087442e48cc

How to cite

Fact over Fiction: Detection of Pathological Hallucinations in Sinhala-to-English Neural Machine Translation — arXiv:2610.11389 — arXiv (en), https://arxiv.org/abs/2610.11389 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6087442e48cc (Metraipolis archive dc_6087442e48cc, sha256 a515d2450559892c68c22652ef6d563477d6c1dff5ff2f1a717b6568f950bfed)
BibTeX
@misc{dc6087442e48cc,
  title = {Fact over Fiction: Detection of Pathological Hallucinations in Sinhala-to-English Neural Machine Translation — arXiv:2610.11389},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11389},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6087442e48cc; sha256 a515d2450559892c68c22652ef6d563477d6c1dff5ff2f1a717b6568f950bfed},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11389. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Neural Machine Translation (NMT) models, while capable of producing highly fluent outputs, remain vulnerable to hallucinations, which are translations that are natural yet semantically unrelated to the source. This vulnerability is acute in low-resource settings like Sinhala-to-English, where weak cross-lingual alignment leads to hallucinations. This paper introduces a framework for reference-free hallucination detection in this language pair. We present a 45,000-sample synthetic dataset generated through a probabilistic chain of five linguistically motivated corruption strategies, with a semantic rescue mechanism that uses character-level similarity to distinguish hallucinations from morphological variants. We fine-tune mDeBERTa-v3 for token-level sequence labelling, reaching a token-level F1 of 0.841 +/- 0.001 over three seeds on a source-disjoint test set, and study a three-signal ensemble integrating neural risk scores, sequence log-probabilities, and cross-lingual semantic embeddings (LaBSE). A source-ablation control shows that the detector relies on the Sinhala source rather than on surface artefacts of the corruption process: shuffling or removing the source reduces sentence-level AUROC from 0.970 to chance. We benchmark eight NMT systems spanning five model families and find that detector firings vary by an order of magnitude across architectures.

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/2610.11389 — 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 2026-10-09T05:47:15.199Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).