<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhao-etal-2025-qspec">
    <titleInfo>
        <title>QSpec: Speculative Decoding with Complementary Quantization Schemes</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Juntao</namePart>
        <namePart type="family">Zhao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Wenhao</namePart>
        <namePart type="family">Lu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Sheng</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Lingpeng</namePart>
        <namePart type="family">Kong</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Chuan</namePart>
        <namePart type="family">Wu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Christos</namePart>
            <namePart type="family">Christodoulopoulos</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Tanmoy</namePart>
            <namePart type="family">Chakraborty</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Carolyn</namePart>
            <namePart type="family">Rose</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Violet</namePart>
            <namePart type="family">Peng</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Suzhou, China</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-332-6</identifier>
    </relatedItem>
    <abstract>Quantization is widely adopted to accelerate inference and reduce memory consumption in large language models (LLMs). While activation-weight joint quantization enables efficient low-precision decoding, it suffers substantial performance degradation on multi-step reasoning tasks. We propose QSPEC, a novel quantization paradigm that decouples efficiency from quality by integrating two complementary schemes via speculative decoding: low-precision joint quantization for fast drafting and high-precision weight-only quantization for accurate verification. QSPEC reuses both weights and KV cache across stages, enabling near-zero-cost switching without retraining or auxiliary models. Compared to high-precision baselines, QSPEC achieves up to 1.64x speedup without quality degradation, and outperforms state-of-the-art speculative decoding methods by up to 1.55x in batched settings. Furthermore, QSPEC supports plug-and-play deployment and generalizes well across model scales, quantization methods, and workloads. These properties make QSPEC a practical and scalable solution for high-fidelity quantized LLM serving under memory-constrained scenarios.</abstract>
    <identifier type="citekey">zhao-etal-2025-qspec</identifier>
    <identifier type="doi">10.18653/v1/2025.emnlp-main.240</identifier>
    <location>
        <url>https://aclanthology.org/2025.emnlp-main.240/</url>
    </location>
    <part>
        <date>2025-11</date>
        <extent unit="page">
            <start>4779</start>
            <end>4795</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T QSpec: Speculative Decoding with Complementary Quantization Schemes
%A Zhao, Juntao
%A Lu, Wenhao
%A Wang, Sheng
%A Kong, Lingpeng
%A Wu, Chuan
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-332-6
%F zhao-etal-2025-qspec
%X Quantization is widely adopted to accelerate inference and reduce memory consumption in large language models (LLMs). While activation-weight joint quantization enables efficient low-precision decoding, it suffers substantial performance degradation on multi-step reasoning tasks. We propose QSPEC, a novel quantization paradigm that decouples efficiency from quality by integrating two complementary schemes via speculative decoding: low-precision joint quantization for fast drafting and high-precision weight-only quantization for accurate verification. QSPEC reuses both weights and KV cache across stages, enabling near-zero-cost switching without retraining or auxiliary models. Compared to high-precision baselines, QSPEC achieves up to 1.64x speedup without quality degradation, and outperforms state-of-the-art speculative decoding methods by up to 1.55x in batched settings. Furthermore, QSPEC supports plug-and-play deployment and generalizes well across model scales, quantization methods, and workloads. These properties make QSPEC a practical and scalable solution for high-fidelity quantized LLM serving under memory-constrained scenarios.
%R 10.18653/v1/2025.emnlp-main.240
%U https://aclanthology.org/2025.emnlp-main.240/
%U https://doi.org/10.18653/v1/2025.emnlp-main.240
%P 4779-4795

Markdown (Informal)

[QSpec: Speculative Decoding with Complementary Quantization Schemes](https://aclanthology.org/2025.emnlp-main.240/) (Zhao et al., EMNLP 2025)

• QSpec: Speculative Decoding with Complementary Quantization Schemes (Zhao et al., EMNLP 2025)

ACL

• Juntao Zhao, Wenhao Lu, Sheng Wang, Lingpeng Kong, and Chuan Wu. 2025. QSpec: Speculative Decoding with Complementary Quantization Schemes. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 4779–4795, Suzhou, China. Association for Computational Linguistics.