<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhou-etal-2026-building">
    <titleInfo>
        <title>Building Long-Term LLM Memory with Neo-Davidsonian Discourse Units</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Sizhe</namePart>
        <namePart type="family">Zhou</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yanzhen</namePart>
        <namePart type="family">Shen</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Siru</namePart>
        <namePart type="family">Ouyang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yuyang</namePart>
        <namePart type="family">Bai</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yiheng</namePart>
        <namePart type="family">Shu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yu</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jiawei</namePart>
        <namePart type="family">Han</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
    </relatedItem>
    <abstract>Long-horizon conversational agents need persistent memory beyond a fixed context window, yet memory effectiveness depends on the units stored and retrieved. Existing systems either discard fine-grained details during compression, retrieve units that are too coarse or context-poor, or fragment discourse into entity-relation triples. We propose Neo-Davidsonian Discourse Units: enriched Elementary Discourse Units (EDUs) that each express one event or fact with normalized entities, temporal grounding, and enough context to stand alone. The resulting units are non-lossy at the session level, atomic for retrieval, and naturally expose event arguments for graph-based associative recall. We instantiate this representation in EMem, which retrieves EDUs with dense representations and relevance filtering based on large language models, and EMem-G, which further organizes sessions, EDUs, and arguments in an event graph and applies Personalized PageRank. On LoCoMo and LongMemEval_S benchmarks, our methods improve over prior state-of-the-art by up to 21.3% while using up to 7\times shorter effective context.</abstract>
    <identifier type="citekey">zhou-etal-2026-building</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.924/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>19494</start>
            <end>19531</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Building Long-Term LLM Memory with Neo-Davidsonian Discourse Units
%A Zhou, Sizhe
%A Shen, Yanzhen
%A Ouyang, Siru
%A Bai, Yuyang
%A Shu, Yiheng
%A Zhang, Yu
%A Han, Jiawei
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F zhou-etal-2026-building
%X Long-horizon conversational agents need persistent memory beyond a fixed context window, yet memory effectiveness depends on the units stored and retrieved. Existing systems either discard fine-grained details during compression, retrieve units that are too coarse or context-poor, or fragment discourse into entity-relation triples. We propose Neo-Davidsonian Discourse Units: enriched Elementary Discourse Units (EDUs) that each express one event or fact with normalized entities, temporal grounding, and enough context to stand alone. The resulting units are non-lossy at the session level, atomic for retrieval, and naturally expose event arguments for graph-based associative recall. We instantiate this representation in EMem, which retrieves EDUs with dense representations and relevance filtering based on large language models, and EMem-G, which further organizes sessions, EDUs, and arguments in an event graph and applies Personalized PageRank. On LoCoMo and LongMemEval_S benchmarks, our methods improve over prior state-of-the-art by up to 21.3% while using up to 7\times shorter effective context.
%U https://aclanthology.org/2026.findings-emnlp.924/
%P 19494-19531

Markdown (Informal)

[Building Long-Term LLM Memory with Neo-Davidsonian Discourse Units](https://aclanthology.org/2026.findings-emnlp.924/) (Zhou et al., Findings 2026)

• Building Long-Term LLM Memory with Neo-Davidsonian Discourse Units (Zhou et al., Findings 2026)

ACL

• Sizhe Zhou, Yanzhen Shen, Siru Ouyang, Yuyang Bai, Yiheng Shu, Yu Zhang, and Jiawei Han. 2026. Building Long-Term LLM Memory with Neo-Davidsonian Discourse Units. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 19494–19531, Budapest, Hungary. Association for Computational Linguistics.