<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="park-hwang-2026-taskpress">
    <titleInfo>
        <title>TaskPress: Query-Agnostic KV Cache Compression via Task-Guided Pruning</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Wonpyo</namePart>
        <namePart type="family">Park</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Seung-Won</namePart>
        <namePart type="family">Hwang</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 inference with large language models (LLMs) is constrained by the linear growth of the key–value (KV) cache to sequence length. While pruning offers mitigation, prevailing methods determine query-specific token importance that cannot be reused across unseen queries. In contrast, we introduce TaskPress, a framework for task-guided, query-agnostic KV cache eviction. Instead of optimizing the cache for a single query, TaskPress constructs a reusable memory representation conditioned on a high-level task guide. The guide functions as a “meta-query” during prefill to filter irrelevant tokens before downstream queries are issued. In addition, TaskPress leverages quantization scale factors as a zero-cost signal for detecting influential representation outliers, providing an efficient proxy for token importance. Experiments on conducted on various tasks with long context input demonstrate that TaskPress efficiently creates a compact, reusable cache across diverse queries.</abstract>
    <identifier type="citekey">park-hwang-2026-taskpress</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.793/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>18423</start>
            <end>18434</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T TaskPress: Query-Agnostic KV Cache Compression via Task-Guided Pruning
%A Park, Wonpyo
%A Hwang, Seung-Won
%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 park-hwang-2026-taskpress
%X Long-context inference with large language models (LLMs) is constrained by the linear growth of the key–value (KV) cache to sequence length. While pruning offers mitigation, prevailing methods determine query-specific token importance that cannot be reused across unseen queries. In contrast, we introduce TaskPress, a framework for task-guided, query-agnostic KV cache eviction. Instead of optimizing the cache for a single query, TaskPress constructs a reusable memory representation conditioned on a high-level task guide. The guide functions as a “meta-query” during prefill to filter irrelevant tokens before downstream queries are issued. In addition, TaskPress leverages quantization scale factors as a zero-cost signal for detecting influential representation outliers, providing an efficient proxy for token importance. Experiments on conducted on various tasks with long context input demonstrate that TaskPress efficiently creates a compact, reusable cache across diverse queries.
%U https://aclanthology.org/2026.emnlp-main.793/
%P 18423-18434

Markdown (Informal)

[TaskPress: Query-Agnostic KV Cache Compression via Task-Guided Pruning](https://aclanthology.org/2026.emnlp-main.793/) (Park & Hwang, EMNLP 2026)

• TaskPress: Query-Agnostic KV Cache Compression via Task-Guided Pruning (Park & Hwang, EMNLP 2026)

ACL

• Wonpyo Park and Seung-Won Hwang. 2026. TaskPress: Query-Agnostic KV Cache Compression via Task-Guided Pruning. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 18423–18434, Budapest, Hungary. Association for Computational Linguistics.