CorefInst：利用大语言模型进行多语言共指消解
CorefInst: Leveraging LLMs for Multilingual Coreference Resolution
AI总结：
本文提出CorefInst，首次利用仅解码器LLM并通过五种指令集建模多语言共指消解，微调后的Llama 3.1在CorefUD v1.2上平均超越SOTA模型2个百分点。
摘要：
共指消解（CR）是自然语言理解中一项关键但具有挑战性的任务，通常受限于特定任务的架构和基于编码器的语言模型，这些模型需要大量训练且缺乏适应性。本研究提出了首个多语言共指消解方法，利用仅解码器的大语言模型（LLM）同时处理显性提及和零提及。文章探讨了如何通过五种不同的指令集，采用受控推理方法为LLM建模共指消解任务。该方法在三个LLM上进行了评估：Llama 3.1、Gemma 2和Mistral 0.3。结果表明，当使用合适的指令集进行指令微调时，LLM可以超越最先进的特定任务架构。具体而言，我们最佳的模型——一个针对多语言共指消解进行完全微调的Llama 3.1——在CorefUD v1.2数据集集合中的所有语言上，平均比领先的多语言共指消解模型（即Corpipe 24单阶段变体）高出2个百分点。
英文摘要：
Coreference Resolution (CR) is a crucial yet challenging task in natural language understanding, often constrained by task-specific architectures and encoder-based language models that demand extensive training and lack adaptability. This study introduces the first multilingual CR methodology which leverages decoder-only LLMs to handle both overt and zero mentions. The article explores how to model the CR task for LLMs via five different instruction sets using a controlled inference method. The approach is evaluated across three LLMs; Llama 3.1, Gemma 2, and Mistral 0.3. The results indicate that LLMs, when instruction-tuned with a suitable instruction set, can surpass state-of-the-art task-specific architectures. Specifically, our best model, a fully fine-tuned Llama 3.1 for multilingual CR, outperforms the leading multilingual CR model (i.e., Corpipe 24 single stage variant) by 2 pp on average across all languages in the CorefUD v1.2 dataset collection.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
