@inproceedings{liu-etal-2025-flexquant,
    title = "{F}lex{Q}uant: A Flexible and Efficient Dynamic Precision Switching Framework for {LLM} Quantization",
    author = "Liu, Fangxin  and
      Wang, Zongwu  and
      Xia, Jinhong  and
      Zhao, Junping  and
      Zhao, Shouren  and
      Li, Jinjin  and
      Liu, Jian  and
      Jiang, Li  and
      Guan, Haibing",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/author-page-yu-wang-polytechnic/2025.findings-emnlp.221/",
    doi = "10.18653/v1/2025.findings-emnlp.221",
    pages = "4152--4161",
    ISBN = "979-8-89176-335-7",
    abstract = "The rapid advancement of large language models (LLMs) has exacerbated the memory bottleneck due to the widening gap between model parameter scaling and hardware capabilities. While post-training quantization techniques effectively reduce memory overhead, existing methods predominantly rely on static quantization strategies, which struggle to adapt to dynamic workloads. To address this, we propose FlexQuant, a dynamic precision-switching framework that optimizes the trade-off between inference speed and accuracy. Leveraging model perplexity entropy and Kullback-Leibler divergence, FlexQuant enables fine-grained, layer-wise mixed-precision quantization and dynamically adjusts bit-widths during each token generation. FlexQuant provides a comprehensive analysis of quantization strategies, introduces a precision requirement model for optimal switching, and implements efficient fine-grained precision management. Evaluations demonstrate that FlexQuant achieves a 1.3{\texttimes} end-to-end speedup across diverse language tasks with negligible accuracy loss introduced. This framework offers a flexible and adaptive solution for efficient LLM deployment."
}