@inproceedings{arora-etal-2025-towards,
    title = "Towards Robust Knowledge Representations in Multilingual {LLM}s for Equivalence and Inheritance based Consistent Reasoning",
    author = "Arora, Gaurav  and
      Merugu, Srujana  and
      Jain, Shreya  and
      Saxena, Vaibhav",
    editor = "Chiruzzo, Luis  and
      Ritter, Alan  and
      Wang, Lu",
    booktitle = "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)",
    month = apr,
    year = "2025",
    address = "Albuquerque, New Mexico",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/fix-sig-urls/2025.naacl-long.394/",
    pages = "7744--7762",
    ISBN = "979-8-89176-189-6",
    abstract = "Reasoning and linguistic skills form the cornerstone of human intelligence, facilitating problem-solving and decision-making. Recent advances in Large Language Models (LLMs) have led to impressive linguistic capabilities and emergent reasoning behaviors, fueling widespread adoption across application domains. However, LLMs still struggle with complex reasoning tasks, highlighting their systemic limitations. In this work, we focus on evaluating whether LLMs have the requisite representations to reason using two foundational relationships: ``equivalence'' and ``inheritance''. We introduce novel tasks and benchmarks spanning six languages and observe that current SOTA LLMs often produce conflicting answers to the same questions across languages in 17.3-57.5{\%} of cases and violate inheritance constraints in up to 37.2{\%} cases. To enhance consistency across languages, we propose novel ``Compositional Representations'' where tokens are represented as composition of equivalent tokens across languages, with resulting conflict reduction (up to -4.7{\%}) indicating benefits of shared LLM representations."
}