When Abundance Conceals Weakness: Knowledge Conflict in Multilingual Models

2026-01-11

arXiv.org

AI Summary

This work addresses the inconsistency between internal beliefs and external evidence in multilingual large language models when confronted with cross-lingual knowledge conflicts—a phenomenon understudied beyond English-centric settings. The authors propose CLEAR, a framework that systematically evaluates how models reconcile language-dependent memorized knowledge with multilingual external evidence across four progressively complex scenarios. They introduce ConflictQA and ConflictingQA, novel multilingual benchmarks spanning ten languages. Their analysis reveals, for the first time, the nuanced roles of language resource abundance and linguistic relatedness: in reasoning-intensive tasks, high-resource languages exert greater influence, whereas in entity-level factual conflicts, linguistic proximity dominates—enabling low-resource but closely related languages to outweigh high-resource yet distantly related ones.

Technology Category

• Natural Language Processing: Machine Translation, Multilinguality, Cross-Lingual NLP
• Knowledge Representation and Reasoning: Knowledge Representation Languages
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Large Language Models (LLMs) encode vast world knowledge across multiple languages, yet their internal beliefs are often unevenly distributed across linguistic spaces. When external evidence contradicts these language-dependent memories, models encounter \emph{cross-lingual knowledge conflict}, a phenomenon largely unexplored beyond English-centric settings. We introduce \textbf{CLEAR}, a \textbf{C}ross-\textbf{L}ingual knowl\textbf{E}dge conflict ev\textbf{A}luation f\textbf{R}amework that systematically examines how multilingual LLMs reconcile conflicting internal beliefs and multilingual external evidence. CLEAR decomposes conflict resolution into four progressive scenarios, from multilingual parametric elicitation to competitive multi-source cross-lingual induction, and systematically evaluates model behavior across two complementary QA benchmarks with distinct task characteristics. We construct multilingual versions of ConflictQA and ConflictingQA covering 10 typologically diverse languages and evaluate six representative LLMs. Our experiments reveal a task-dependent decision dichotomy. In reasoning-intensive tasks, conflict resolution is dominated by language resource abundance, with high-resource languages exerting stronger persuasive power. In contrast, for entity-centric factual conflicts, linguistic affinity, not resource scale, becomes decisive, allowing low-resource but linguistically aligned languages to outperform distant high-resource ones.

Problem

• cross-lingual knowledge conflict
• multilingual LLMs
• knowledge conflict
• language resource abundance
• linguistic affinity

Innovation

• cross-lingual knowledge conflict
• multilingual LLMs
• CLEAR framework
• language resource abundance
• linguistic affinity