<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="huang-etal-2026-kv">
    <titleInfo>
        <title>KV Admission: Learning What to Write for Efficient Long-Context LLM Inference</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yen-Chieh</namePart>
        <namePart type="family">Huang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Pi-Cheng</namePart>
        <namePart type="family">Hsiu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Rui</namePart>
        <namePart type="family">Fang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ming-syan</namePart>
        <namePart type="family">Chen</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>Long-context LLM inference is bottlenecked by the quadratic attention complexity and linear Key-Value (KV) cache growth. Prior approaches mitigate this via post-hoc selection or eviction but overlook the root inefficiency: *indiscriminate token admission*. In this paper, we formalize KV management as a causal system of three primitives: KV Admission, Selection, and Eviction. We instantiate KV Admission via Write-Gated KV (WG-KV), a lightweight mechanism that learns to predict token utility *before* cache entry. By filtering out redundant states early to maintain a compact global cache alongside a sliding local cache, WG-KV significantly reduces memory usage and accelerates both prefill and decode phases. Our results demonstrate that *learning what to write* is a principled and practical recipe for efficient long-context inference. Code is available at https://github.com/EMCLab-Sinica/WG-KV.</abstract>
    <identifier type="citekey">huang-etal-2026-kv</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.1063/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>24411</start>
            <end>24439</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T KV Admission: Learning What to Write for Efficient Long-Context LLM Inference
%A Huang, Yen-Chieh
%A Hsiu, Pi-Cheng
%A Fang, Rui
%A Chen, Ming-syan
%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 huang-etal-2026-kv
%X Long-context LLM inference is bottlenecked by the quadratic attention complexity and linear Key-Value (KV) cache growth. Prior approaches mitigate this via post-hoc selection or eviction but overlook the root inefficiency: *indiscriminate token admission*. In this paper, we formalize KV management as a causal system of three primitives: KV Admission, Selection, and Eviction. We instantiate KV Admission via Write-Gated KV (WG-KV), a lightweight mechanism that learns to predict token utility *before* cache entry. By filtering out redundant states early to maintain a compact global cache alongside a sliding local cache, WG-KV significantly reduces memory usage and accelerates both prefill and decode phases. Our results demonstrate that *learning what to write* is a principled and practical recipe for efficient long-context inference. Code is available at https://github.com/EMCLab-Sinica/WG-KV.
%U https://aclanthology.org/2026.emnlp-main.1063/
%P 24411-24439

Markdown (Informal)

[KV Admission: Learning What to Write for Efficient Long-Context LLM Inference](https://aclanthology.org/2026.emnlp-main.1063/) (Huang et al., EMNLP 2026)

• KV Admission: Learning What to Write for Efficient Long-Context LLM Inference (Huang et al., EMNLP 2026)

ACL

• Yen-Chieh Huang, Pi-Cheng Hsiu, Rui Fang, and Ming-syan Chen. 2026. KV Admission: Learning What to Write for Efficient Long-Context LLM Inference. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 24411–24439, Budapest, Hungary. Association for Computational Linguistics.