arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.01116cs.CL
收藏

基于对比学习的逻辑导向检索增强

Logic-Oriented Retriever Enhancement via Contrastive Learning

Shanghai Institute of Artificial Intelligence for Education, East China Normal University, China(上海人工智能教育研究院,华东师范大学);Education Technology, East China Normal University, China(教育技术,华东师范大学)

Wenxuan Zhang, Yuan-Hao Jiang, Changyong Qi, Rui Jia, Yonghe Wu

AI总结:

LORE通过对比学习提升检索器的逻辑分析能力,无需额外资源,有效增强检索和生成性能。

摘要:

大型语言模型(LLMs)在知识密集型任务中表现不佳,因为检索器往往过度依赖表面相似性,而在涉及复杂逻辑关系的查询中表现不佳。模型表示中具有逻辑分析能力,但这种能力在标准训练中仍未被充分利用。LORE(逻辑导向检索增强)引入了细粒度对比学习,以激活这种潜在能力,引导嵌入向与逻辑结构一致的证据方向发展,而不是浅层相似性。LORE不需要外部监督、资源或预检索分析,保持索引兼容性,并在保持效率的同时,一致提高检索效用和下游生成能力。数据集和代码可在https://github.com/mazehart/Lore-RAG上公开获取。

英文摘要:

Large language models (LLMs) struggle in knowledge-intensive tasks, as retrievers often overfit to surface similarity and fail on queries involving complex logical relations. The capacity for logical analysis is inherent in model representations but remains underutilized in standard training. LORE (Logic ORiented Retriever Enhancement) introduces fine-grained contrastive learning to activate this latent capacity, guiding embeddings toward evidence aligned with logical structure rather than shallow similarity. LORE requires no external upervision, resources, or pre-retrieval analysis, remains index-compatible, and consistently improves retrieval utility and downstream generation while maintaining efficiency. The datasets and code are publicly available at https://github.com/mazehart/Lore-RAG.

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。

↑