<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="peng-etal-2026-quanag">
    <titleInfo>
        <title>QuaNAG: LLM-driven Memory-bounded Neural Architecture Generation via Quantization-branched Exploration</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yingqi</namePart>
        <namePart type="family">Peng</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yifan</namePart>
        <namePart type="family">Lin</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kaijie</namePart>
        <namePart type="family">Gong</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yi</namePart>
        <namePart type="family">Gao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Wei</namePart>
        <namePart type="family">Dong</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-459-0</identifier>
    </relatedItem>
    <abstract>Human Activity Recognition (HAR) on memory-constrained edge devices demands heterogeneous, accurate, resource-adaptive architectures that are costly to design manually. Neural Architecture Search (NAS) provides an automated solution for model design, and its integration with quantization can further expose high-quality architectures under edge memory budgets. Recent NAS increasingly adopts LLM-based methods, including prompt-engineering and fine-tuning approaches. Prompt-engineering approaches suffer from 4-10 hours generation time, while fine-tuning approaches achieve minute-level generation but often sacrifice architecture quality due to the complexity of quantization. We propose QuaNAG, a Neural Architecture Generation (NAG) framework built on the insight that quantization is not an ordinary architecture attribute: it can change model feasibility, alter accuracy, and expand the feasible search space. QuaNAG therefore models quantization as branch-level search decisions and transfers quantization-aware search knowledge into an LLM through curriculum learning. Experiments on demanding HAR benchmarks over five baselines, including in-domain evaluations, out-of-domain evaluations, and on-device experiments, demonstrate that QuaNAG achieves 8.93% average accuracy improvement with several minutes generation time, satisfying all deployment requirements. Our code and model are available at https://github.com/pyq623/QuaNAG.</abstract>
    <identifier type="citekey">peng-etal-2026-quanag</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.2205/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>50143</start>
            <end>50174</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T QuaNAG: LLM-driven Memory-bounded Neural Architecture Generation via Quantization-branched Exploration
%A Peng, Yingqi
%A Lin, Yifan
%A Gong, Kaijie
%A Gao, Yi
%A Dong, Wei
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F peng-etal-2026-quanag
%X Human Activity Recognition (HAR) on memory-constrained edge devices demands heterogeneous, accurate, resource-adaptive architectures that are costly to design manually. Neural Architecture Search (NAS) provides an automated solution for model design, and its integration with quantization can further expose high-quality architectures under edge memory budgets. Recent NAS increasingly adopts LLM-based methods, including prompt-engineering and fine-tuning approaches. Prompt-engineering approaches suffer from 4-10 hours generation time, while fine-tuning approaches achieve minute-level generation but often sacrifice architecture quality due to the complexity of quantization. We propose QuaNAG, a Neural Architecture Generation (NAG) framework built on the insight that quantization is not an ordinary architecture attribute: it can change model feasibility, alter accuracy, and expand the feasible search space. QuaNAG therefore models quantization as branch-level search decisions and transfers quantization-aware search knowledge into an LLM through curriculum learning. Experiments on demanding HAR benchmarks over five baselines, including in-domain evaluations, out-of-domain evaluations, and on-device experiments, demonstrate that QuaNAG achieves 8.93% average accuracy improvement with several minutes generation time, satisfying all deployment requirements. Our code and model are available at https://github.com/pyq623/QuaNAG.
%U https://aclanthology.org/2026.emnlp-main.2205/
%P 50143-50174

Markdown (Informal)

[QuaNAG: LLM-driven Memory-bounded Neural Architecture Generation via Quantization-branched Exploration](https://aclanthology.org/2026.emnlp-main.2205/) (Peng et al., EMNLP 2026)

• QuaNAG: LLM-driven Memory-bounded Neural Architecture Generation via Quantization-branched Exploration (Peng et al., EMNLP 2026)

ACL

• Yingqi Peng, Yifan Lin, Kaijie Gong, Yi Gao, and Wei Dong. 2026. QuaNAG: LLM-driven Memory-bounded Neural Architecture Generation via Quantization-branched Exploration. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 50143–50174, Budapest, Hungary. Association for Computational Linguistics.