用翻译提升多语言幻觉检测，零训练即获高分
AILS-NTUA at SemEval-2025 Task 3: Leveraging Large Language Models and Translation Strategies for Multilingual Hallucination Detection
- 将多语言文本转译成英文后用大模型检测幻觉
- 在低资源语言中拿下两个第一，表现稳定
- 适合想快速部署多语言幻觉检测的团队
多语言幻觉检测仍是一个研究不足的挑战，本次穆-舒姆共享任务旨在推动该方向发展。本文提出一种高效、无需训练的大语言模型提示策略，通过将多语言文本片段翻译为英语来增强检测能力。该方法在多个语言上取得具有竞争力的排名，在低资源语言中获得两个第一名。结果的一致性表明，该翻译策略在不同源语言下均有效，具备广泛适用性。
原文摘要 · Abstract (English)
Multilingual hallucination detection stands as an underexplored challenge, which the Mu-SHROOM shared task seeks to address. In this work, we propose an efficient, training-free LLM prompting strategy that enhances detection by translating multilingual text spans into English. Our approach achieves competitive rankings across multiple languages, securing two first positions in low-resource languages. The consistency of our results highlights the effectiveness of our translation strategy for hallucination detection, demonstrating its applicability regardless of the source language.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。