@inproceedings{wang-etal-2025-tracing,
    title = "Tracing and Dissecting How {LLM}s Recall Factual Knowledge for Real World Questions",
    author = "Wang, Yiqun  and
      Wan, Chaoqun  and
      Hu, Sile  and
      Zhang, Yonggang  and
      Tian, Xiang  and
      Chen, Yaowu  and
      Shen, Xu  and
      Ye, Jieping",
    editor = "Che, Wanxiang  and
      Nabende, Joyce  and
      Shutova, Ekaterina  and
      Pilehvar, Mohammad Taher",
    booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingestion-acl-25/2025.acl-long.1133/",
    pages = "23246--23271",
    ISBN = "979-8-89176-251-0",
    abstract = "Recent advancements in large language models (LLMs) have shown promising ability to perform commonsense reasoning, bringing machines closer to human-like understanding. However, deciphering the internal reasoning processes of LLMs remains challenging due to the complex interdependencies among generated tokens, especially in practical question-answering. In this study, we introduce a two-dimensional analysis framework{---}comprising token back-tracing and individual token decoding{---}to uncover how LLMs conduct factual knowledge recall. Through explanatory analysis of three typical reasoning datasets, we identify a consistent three-phase pattern: Subject Augmentation and Broadcasting, Object Retrieval and Reranking, and Conclusion Fusion and Generation. Our findings reveal that LLMs do not lack relevant knowledge but struggle to select the most accurate information based on context during the retrieval and rerank phase. Leveraging these findings, we apply representation engineering and selective fine-tuning to target specific modules responsible for retrieval and rerank errors. Experimental results show large improvements in response accuracy for both in-domain and out-of-domain settings, validating the rationality of the interpreting result."
}