arXiv 2601.029172026-01-06cs.CLcs.AI

RAL2M：检索增强的学习-匹配以对抗幻觉在合规保障服务系统中

RAL2M: Retrieval Augmented Learning-To-Match Against Hallucination in Compliance-Guaranteed Service Systems

Hong Kong Polytechnic University（香港理工大学）；New York University Shanghai（纽约大学上海分校）；Hong Kong University of Science and Technology（香港科技大学）；Beijing University of Posts and Telecommunications（北京邮电大学）

Mengze Hong, Di Jiang, Jiangtao Wen, Zhiyang Su, Yawen Li, Yanjie Sun, Guan Wang, Chen Jason Zhang

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

RAL2M通过检索增强的学习-匹配框架，利用众智消除生成幻觉，提升合规保障服务系统的响应可靠性。

摘要：

幻觉是LLM驱动的服务系统中的主要问题，需要显式知识支撑以确保响应的合规性。在本文中，我们介绍了检索增强的学习-匹配（RAL2M）框架，该框架通过将LLM重新定位为查询-响应匹配的裁判器，在基于检索的系统中消除生成幻觉，提供了一种比纯生成方法更稳健的替代方案。为进一步减轻判断幻觉，我们提出了一种查询自适应的潜在集合策略，该策略显式建模异构模型能力和LLM之间的互相关性，推导出校准的共识决策。在大规模基准测试中进行的广泛实验表明，所提出的方法有效利用了“众智”的智慧，并显著优于强大的基线方法。最后，我们讨论了未来工作中进一步利用潜在表示的最佳实践和有前景的方向。

英文摘要：

Hallucination is a major concern in LLM-driven service systems, necessitating explicit knowledge grounding for compliance-guaranteed responses. In this paper, we introduce Retrieval-Augmented Learning-to-Match (RAL2M), a novel framework that eliminates generation hallucination by repositioning LLMs as query-response matching judges within a retrieval-based system, providing a robust alternative to purely generative approaches. To further mitigate judgment hallucination, we propose a query-adaptive latent ensemble strategy that explicitly models heterogeneous model competence and interdependencies among LLMs, deriving a calibrated consensus decision. Extensive experiments on large-scale benchmarks demonstrate that the proposed method effectively leverages the "wisdom of the crowd" and significantly outperforms strong baselines. Finally, we discuss best practices and promising directions for further exploiting latent representations in future work.

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。