<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-training">
    <titleInfo>
        <title>Training-Free Adaptive Speculative Decoding via Linguistic Priors</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Jingyi</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jiaqi</namePart>
        <namePart type="family">Huang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Zunnan</namePart>
        <namePart type="family">Xu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jun</namePart>
        <namePart type="family">Zhou</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kehong</namePart>
        <namePart type="family">Yuan</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xiang</namePart>
        <namePart type="family">Qian</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: ACL 2026</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Maria</namePart>
            <namePart type="family">Liakata</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Viviane</namePart>
            <namePart type="given">P</namePart>
            <namePart type="family">Moreira</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jiajun</namePart>
            <namePart type="family">Zhang</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">David</namePart>
            <namePart type="family">Jurgens</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">San Diego, California, United States</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-395-1</identifier>
    </relatedItem>
    <abstract>Speculative decoding (SPD) has emerged as a promising technique to accelerate Large Language Model (LLM) inference. However, current approaches typically enforce a uniform verification standard, neglecting the inherent heterogeneity of natural language and failing to distinguish between semantically-rich content and structurally-predictable syntax. In this paper, we propose LinguaSpec, a training-free framework that leverages linguistic priors to enable adaptive drafting and verification. Specifically, we introduce: (1) a Static Linguistic Probe (SLP) to categorize tokens with zero latency; (2) Syntactic Normalized Surprisal (SNS) to calibrate uncertainty against category-specific entropy; and (3) a dual strategy of Syntactically-Guided Elastic Expansion and POS-Adaptive Deferred Verification to dynamically adjust drafting depth and verification rigor. By balancing semantic integrity with structural efficiency, LinguaSpec significantly accelerates inference without requiring additional training. Experimental results demonstrate its superior performance across diverse benchmarks.</abstract>
    <identifier type="citekey">wang-etal-2026-training</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-acl.1065/</url>
    </location>
    <part>
        <date>2026-07</date>
        <extent unit="page">
            <start>21184</start>
            <end>21194</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Training-Free Adaptive Speculative Decoding via Linguistic Priors
%A Wang, Jingyi
%A Huang, Jiaqi
%A Xu, Zunnan
%A Zhou, Jun
%A Yuan, Kehong
%A Qian, Xiang
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F wang-etal-2026-training
%X Speculative decoding (SPD) has emerged as a promising technique to accelerate Large Language Model (LLM) inference. However, current approaches typically enforce a uniform verification standard, neglecting the inherent heterogeneity of natural language and failing to distinguish between semantically-rich content and structurally-predictable syntax. In this paper, we propose LinguaSpec, a training-free framework that leverages linguistic priors to enable adaptive drafting and verification. Specifically, we introduce: (1) a Static Linguistic Probe (SLP) to categorize tokens with zero latency; (2) Syntactic Normalized Surprisal (SNS) to calibrate uncertainty against category-specific entropy; and (3) a dual strategy of Syntactically-Guided Elastic Expansion and POS-Adaptive Deferred Verification to dynamically adjust drafting depth and verification rigor. By balancing semantic integrity with structural efficiency, LinguaSpec significantly accelerates inference without requiring additional training. Experimental results demonstrate its superior performance across diverse benchmarks.
%U https://aclanthology.org/2026.findings-acl.1065/
%P 21184-21194

Markdown (Informal)

[Training-Free Adaptive Speculative Decoding via Linguistic Priors](https://aclanthology.org/2026.findings-acl.1065/) (Wang et al., Findings 2026)

• Training-Free Adaptive Speculative Decoding via Linguistic Priors (Wang et al., Findings 2026)

ACL

• Jingyi Wang, Jiaqi Huang, Zunnan Xu, Jun Zhou, Kehong Yuan, and Xiang Qian. 2026. Training-Free Adaptive Speculative Decoding via Linguistic Priors. In Findings of the Association for Computational Linguistics: ACL 2026, pages 21184–21194, San Diego, California, United States. Association for Computational Linguistics.