<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="fei-etal-2026-memory">
    <titleInfo>
        <title>Memory Beyond Recall: A Dual-Process Cognitive Memory System for Self-Evolving LLM Agents</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Tianxiang</namePart>
        <namePart type="family">Fei</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mingyang</namePart>
        <namePart type="family">Song</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mao</namePart>
        <namePart type="family">Zheng</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xiang</namePart>
        <namePart type="family">Yu</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-term memory for an LLM agent requires more than retrieving a relevant passage. Many existing systems address belief revision, causal connections and cross-domain abstraction primarily through storage and retrieval mechanisms, limiting their ability to support implicit personalisation as a user’s preferences evolve. We present DPCM, which organises memory into a capability hierarchy spanning raw inputs, atomic facts, belief trajectories, identity, domain schemas, prospective intentions and cross-domain patterns. A synchronous daytime writer (SYSTEM 1) records revisions as doubly linked supersedes chains; an asynchronous nighttime engine (SYSTEM 2) induces schemas and intentions, then uses cross-domain collisions to derive core schemas. Dual-process theory motivates this separation between online encoding and offline abstraction without implying that agent memory reproduces human cognition. Across single-run evaluations on LongMemEval, PersonaMem and PersonaMem-v2, SYSTEM 2 yields its largest observed gain on implicit cross-session inference (+5.20 on PersonaMem-v2) and its smallest on the recall-oriented LongMemEval aggregate (+0.33). When the inference budget is limited to 8k tokens and the full history no longer fits, DPCMfull exceeds the long-context baseline by 10.5–17.9 points across both backbones.</abstract>
    <identifier type="citekey">fei-etal-2026-memory</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.1918/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>40897</start>
            <end>40914</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Memory Beyond Recall: A Dual-Process Cognitive Memory System for Self-Evolving LLM Agents
%A Fei, Tianxiang
%A Song, Mingyang
%A Zheng, Mao
%A Yu, Xiang
%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 fei-etal-2026-memory
%X Long-term memory for an LLM agent requires more than retrieving a relevant passage. Many existing systems address belief revision, causal connections and cross-domain abstraction primarily through storage and retrieval mechanisms, limiting their ability to support implicit personalisation as a user’s preferences evolve. We present DPCM, which organises memory into a capability hierarchy spanning raw inputs, atomic facts, belief trajectories, identity, domain schemas, prospective intentions and cross-domain patterns. A synchronous daytime writer (SYSTEM 1) records revisions as doubly linked supersedes chains; an asynchronous nighttime engine (SYSTEM 2) induces schemas and intentions, then uses cross-domain collisions to derive core schemas. Dual-process theory motivates this separation between online encoding and offline abstraction without implying that agent memory reproduces human cognition. Across single-run evaluations on LongMemEval, PersonaMem and PersonaMem-v2, SYSTEM 2 yields its largest observed gain on implicit cross-session inference (+5.20 on PersonaMem-v2) and its smallest on the recall-oriented LongMemEval aggregate (+0.33). When the inference budget is limited to 8k tokens and the full history no longer fits, DPCMfull exceeds the long-context baseline by 10.5–17.9 points across both backbones.
%U https://aclanthology.org/2026.findings-emnlp.1918/
%P 40897-40914

Markdown (Informal)

[Memory Beyond Recall: A Dual-Process Cognitive Memory System for Self-Evolving LLM Agents](https://aclanthology.org/2026.findings-emnlp.1918/) (Fei et al., Findings 2026)

• Memory Beyond Recall: A Dual-Process Cognitive Memory System for Self-Evolving LLM Agents (Fei et al., Findings 2026)

ACL

• Tianxiang Fei, Mingyang Song, Mao Zheng, and Xiang Yu. 2026. Memory Beyond Recall: A Dual-Process Cognitive Memory System for Self-Evolving LLM Agents. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 40897–40914, Budapest, Hungary. Association for Computational Linguistics.