Do LLM hallucination detectors suffer from low-resource effect?
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909999011201024 |
|---|---|
| author | Datta, Debtanu Chilukuri, Mohan Kishore Kumar, Yash Ghosh, Saptarshi Zafar, Muhammad Bilal |
| author_facet | Datta, Debtanu Chilukuri, Mohan Kishore Kumar, Yash Ghosh, Saptarshi Zafar, Muhammad Bilal |
| contents | LLMs, while outperforming humans in a wide range of tasks, can still fail in unanticipated ways. We focus on two pervasive failure modes: (i) hallucinations, where models produce incorrect information about the world, and (ii) the low-resource effect, where the models show impressive performance in high-resource languages like English but the performance degrades significantly in low-resource languages like Bengali. We study the intersection of these issues and ask: do hallucination detectors suffer from the low-resource effect? We conduct experiments on five tasks across three domains (factual recall, STEM, and Humanities). Experiments with four LLMs and three hallucination detectors reveal a curious finding: As expected, the task accuracies in low-resource languages experience large drops (compared to English). However, the drop in detectors' accuracy is often several times smaller than the drop in task accuracy. Our findings suggest that even in low-resource languages, the internal mechanisms of LLMs might encode signals about their uncertainty. Further, the detectors are robust within language (even for non-English) and in multilingual setups, but not in cross-lingual settings without in-language supervision. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_16766 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Do LLM hallucination detectors suffer from low-resource effect? Datta, Debtanu Chilukuri, Mohan Kishore Kumar, Yash Ghosh, Saptarshi Zafar, Muhammad Bilal Computation and Language Artificial Intelligence LLMs, while outperforming humans in a wide range of tasks, can still fail in unanticipated ways. We focus on two pervasive failure modes: (i) hallucinations, where models produce incorrect information about the world, and (ii) the low-resource effect, where the models show impressive performance in high-resource languages like English but the performance degrades significantly in low-resource languages like Bengali. We study the intersection of these issues and ask: do hallucination detectors suffer from the low-resource effect? We conduct experiments on five tasks across three domains (factual recall, STEM, and Humanities). Experiments with four LLMs and three hallucination detectors reveal a curious finding: As expected, the task accuracies in low-resource languages experience large drops (compared to English). However, the drop in detectors' accuracy is often several times smaller than the drop in task accuracy. Our findings suggest that even in low-resource languages, the internal mechanisms of LLMs might encode signals about their uncertainty. Further, the detectors are robust within language (even for non-English) and in multilingual setups, but not in cross-lingual settings without in-language supervision. |
| title | Do LLM hallucination detectors suffer from low-resource effect? |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2601.16766 |