NCL-UoR at SemEval-2025 Task 3: Detecting Multilingual Hallucination and Related Observable Overgeneration Text Spans with Modified RefChecker and Modified SeflCheckGPT

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hong, Jiaying, Markchom, Thanet, Xu, Jianfei, Wu, Tong, Liang, Huizhi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915281953095680
author Hong, Jiaying
Markchom, Thanet
Xu, Jianfei
Wu, Tong
Liang, Huizhi
author_facet Hong, Jiaying
Markchom, Thanet
Xu, Jianfei
Wu, Tong
Liang, Huizhi
contents SemEval-2025 Task 3 (Mu-SHROOM) focuses on detecting hallucinations in content generated by various large language models (LLMs) across multiple languages. This task involves not only identifying the presence of hallucinations but also pinpointing their specific occurrences. To tackle this challenge, this study introduces two methods: modified RefChecker and modified SelfCheckGPT. The modified RefChecker integrates prompt-based factual verification into References, structuring them as claim-based tests rather than single external knowledge sources. The modified SelfCheckGPT incorporates external knowledge to overcome its reliance on internal knowledge. In addition, both methods' original prompt designs are enhanced to identify hallucinated words within LLM-generated texts. Experimental results demonstrate the effectiveness of the approach, achieving a high ranking on the test dataset in detecting hallucinations across various languages, with an average IoU of 0.5310 and an average COR of 0.5669.
format Preprint
id arxiv_https___arxiv_org_abs_2503_01921
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle NCL-UoR at SemEval-2025 Task 3: Detecting Multilingual Hallucination and Related Observable Overgeneration Text Spans with Modified RefChecker and Modified SeflCheckGPT
Hong, Jiaying
Markchom, Thanet
Xu, Jianfei
Wu, Tong
Liang, Huizhi
Computation and Language
Artificial Intelligence
SemEval-2025 Task 3 (Mu-SHROOM) focuses on detecting hallucinations in content generated by various large language models (LLMs) across multiple languages. This task involves not only identifying the presence of hallucinations but also pinpointing their specific occurrences. To tackle this challenge, this study introduces two methods: modified RefChecker and modified SelfCheckGPT. The modified RefChecker integrates prompt-based factual verification into References, structuring them as claim-based tests rather than single external knowledge sources. The modified SelfCheckGPT incorporates external knowledge to overcome its reliance on internal knowledge. In addition, both methods' original prompt designs are enhanced to identify hallucinated words within LLM-generated texts. Experimental results demonstrate the effectiveness of the approach, achieving a high ranking on the test dataset in detecting hallucinations across various languages, with an average IoU of 0.5310 and an average COR of 0.5669.
title NCL-UoR at SemEval-2025 Task 3: Detecting Multilingual Hallucination and Related Observable Overgeneration Text Spans with Modified RefChecker and Modified SeflCheckGPT
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2503.01921