Span-Level Hallucination Detection for LLM-Generated Answers
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909594735869952 |
|---|---|
| author | Elchafei, Passant Abu-Elkheir, Mervet |
| author_facet | Elchafei, Passant Abu-Elkheir, Mervet |
| contents | Detecting spans of hallucination in LLM-generated answers is crucial for improving factual consistency. This paper presents a span-level hallucination detection framework for the SemEval-2025 Shared Task, focusing on English and Arabic texts. Our approach integrates Semantic Role Labeling (SRL) to decompose the answer into atomic roles, which are then compared with a retrieved reference context obtained via question-based LLM prompting. Using a DeBERTa-based textual entailment model, we evaluate each role semantic alignment with the retrieved context. The entailment scores are further refined through token-level confidence measures derived from output logits, and the combined scores are used to detect hallucinated spans. Experiments on the Mu-SHROOM dataset demonstrate competitive performance. Additionally, hallucinated spans have been verified through fact-checking by prompting GPT-4 and LLaMA. Our findings contribute to improving hallucination detection in LLM-generated responses. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_18639 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Span-Level Hallucination Detection for LLM-Generated Answers Elchafei, Passant Abu-Elkheir, Mervet Computation and Language Detecting spans of hallucination in LLM-generated answers is crucial for improving factual consistency. This paper presents a span-level hallucination detection framework for the SemEval-2025 Shared Task, focusing on English and Arabic texts. Our approach integrates Semantic Role Labeling (SRL) to decompose the answer into atomic roles, which are then compared with a retrieved reference context obtained via question-based LLM prompting. Using a DeBERTa-based textual entailment model, we evaluate each role semantic alignment with the retrieved context. The entailment scores are further refined through token-level confidence measures derived from output logits, and the combined scores are used to detect hallucinated spans. Experiments on the Mu-SHROOM dataset demonstrate competitive performance. Additionally, hallucinated spans have been verified through fact-checking by prompting GPT-4 and LLaMA. Our findings contribute to improving hallucination detection in LLM-generated responses. |
| title | Span-Level Hallucination Detection for LLM-Generated Answers |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2504.18639 |