@inproceedings{tatarinov-etal-2026-kg,
    title = "{KG}-{M}u{LQA}: A Framework for {KG}-based Multi-Level {QA} Extraction and Long-Context {LLM} Evaluation",
    author = "Tatarinov, Nikita  and
      Kannan, Vidhyakshaya  and
      Srinivasa, Haricharana  and
      Raj, Arnav  and
      Anand, Harpreet Singh  and
      Singh, Varun  and
      Luthra, Aditya  and
      Lade, Ravij  and
      Shah, Agam  and
      Chava, Sudheer",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.151/",
    pages = "3323--3359",
    ISBN = "979-8-89176-390-6",
    abstract = "We introduce KG-MuLQA (Knowledge-Graph-based Multi-Level Question-Answer Extraction): a framework that (1) extracts QA pairs at multiple complexity levels (2) along three key dimensions {--} multi-hop retrieval, set operations, and answer plurality, (3) by leveraging knowledge-graph-based document representations. This approach enables fine-grained assessment of model performance across controlled difficulty levels. Using this framework, we construct a dataset of 20,139 QA pairs based on financial credit agreements and evaluate 16 proprietary and open-weight Large Language Models, observing that even the best-performing models struggle with set-based comparisons and multi-hop reasoning over long contexts. Our analysis reveals systematic failure modes tied to semantic misinterpretation and inability to handle implicit relations."
}