Abstract
This paper describes our submission for SemEval-2024 Task 6: SHROOM, a Shared-task on Hallucinations and Related Observable Overgeneration Mistakes. We propose four groups of methods for hallucination detection: 1) Entailment Recognition; 2) Similarity Search; 3) Factuality Verification; 4) Confidence Estimation. The four methods rely on either the semantic relationship between the hypothesis and its source (target) or on the model-aware features during decoding. We participated in both the model-agnostic and model-aware tracks. Our method’s effectiveness is validated by our high rankings 3rd in the model-agnostic track and 5th in the model-aware track. We have released our code on GitHub.- Anthology ID:
- 2024.semeval-1.253
- Volume:
- Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024)
- Month:
- June
- Year:
- 2024
- Address:
- Mexico City, Mexico
- Editors:
- Atul Kr. Ojha, A. Seza Doğruöz, Harish Tayyar Madabushi, Giovanni Da San Martino, Sara Rosenthal, Aiala Rosá
- Venue:
- SemEval
- SIG:
- SIGLEX
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 1788–1797
- Language:
- URL:
- https://preview.aclanthology.org/add_missing_videos/2024.semeval-1.253/
- DOI:
- 10.18653/v1/2024.semeval-1.253
- Cite (ACL):
- Wei Liu, Wanyao Shi, Zijian Zhang, and Hui Huang. 2024. HIT-MI&T Lab at SemEval-2024 Task 6: DeBERTa-based Entailment Model is a Reliable Hallucination Detector. In Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024), pages 1788–1797, Mexico City, Mexico. Association for Computational Linguistics.
- Cite (Informal):
- HIT-MI&T Lab at SemEval-2024 Task 6: DeBERTa-based Entailment Model is a Reliable Hallucination Detector (Liu et al., SemEval 2024)
- PDF:
- https://preview.aclanthology.org/add_missing_videos/2024.semeval-1.253.pdf