@inproceedings{christou-etal-2026-beyond,
title = "Beyond Retrieval: Deterministic Memory Governance with Pairwise Constraints and Certified Repair for Long-Term Language Agents",
author = "Christou, Panayiotis and
Palma, Miguel and
Du, Zefan and
Hua, Kaixun and
Mao, Ying",
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the {A}sia-Pacific Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.aacl-long.246/",
pages = "5501--5518",
ISBN = "979-8-89176-439-2",
abstract = "Long-term memory significantly enhances language-model agents but introduces a critical failure mode: retrieval surfaces relevant yet stale, deleted, or mutually inconsistent facts. A query about a current address may retrieve the correct address alongside deleted ones. The answer is present, but the memory set is unsafe. We reframe agent memory as a governed object-selection problem and present MemGov, a substrate-agnostic middleware layer between existing retrieval pipelines and agent context. MemGov wraps retrieved evidence as explicit memory objects, detects must-link and cannot-link constraints, and applies deterministic constraint-aware repair, returning either a certified safe memory set or an auditable obstruction certificate rather than silently passing unsafe combinations. Evaluations on controlled governance scenarios and MemoryAgentBench-style end-to-end tasks, deliberately separating retrieval quality from governance decisions, demonstrate that while raw retrieval recovers answer-bearing evidence successfully, it simultaneously returns detected unsafe memory combinations in 47-51{\%} of cases. The same retrieval substrates governed by MemGov reduce unsafe merges to 0{\%} while preserving answer utility within about 2-4 percentage points, with fully auditable reasoning for all repair and refusal decisions. Additional diagnostics assess eligibility, temporal constraints, and paired answer retention."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="christou-etal-2026-beyond">
<titleInfo>
<title>Beyond Retrieval: Deterministic Memory Governance with Pairwise Constraints and Certified Repair for Long-Term Language Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Panayiotis</namePart>
<namePart type="family">Christou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Miguel</namePart>
<namePart type="family">Palma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zefan</namePart>
<namePart type="family">Du</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kaixun</namePart>
<namePart type="family">Hua</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ying</namePart>
<namePart type="family">Mao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-439-2</identifier>
</relatedItem>
<abstract>Long-term memory significantly enhances language-model agents but introduces a critical failure mode: retrieval surfaces relevant yet stale, deleted, or mutually inconsistent facts. A query about a current address may retrieve the correct address alongside deleted ones. The answer is present, but the memory set is unsafe. We reframe agent memory as a governed object-selection problem and present MemGov, a substrate-agnostic middleware layer between existing retrieval pipelines and agent context. MemGov wraps retrieved evidence as explicit memory objects, detects must-link and cannot-link constraints, and applies deterministic constraint-aware repair, returning either a certified safe memory set or an auditable obstruction certificate rather than silently passing unsafe combinations. Evaluations on controlled governance scenarios and MemoryAgentBench-style end-to-end tasks, deliberately separating retrieval quality from governance decisions, demonstrate that while raw retrieval recovers answer-bearing evidence successfully, it simultaneously returns detected unsafe memory combinations in 47-51% of cases. The same retrieval substrates governed by MemGov reduce unsafe merges to 0% while preserving answer utility within about 2-4 percentage points, with fully auditable reasoning for all repair and refusal decisions. Additional diagnostics assess eligibility, temporal constraints, and paired answer retention.</abstract>
<identifier type="citekey">christou-etal-2026-beyond</identifier>
<location>
<url>https://aclanthology.org/2026.aacl-long.246/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>5501</start>
<end>5518</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Beyond Retrieval: Deterministic Memory Governance with Pairwise Constraints and Certified Repair for Long-Term Language Agents
%A Christou, Panayiotis
%A Palma, Miguel
%A Du, Zefan
%A Hua, Kaixun
%A Mao, Ying
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-439-2
%F christou-etal-2026-beyond
%X Long-term memory significantly enhances language-model agents but introduces a critical failure mode: retrieval surfaces relevant yet stale, deleted, or mutually inconsistent facts. A query about a current address may retrieve the correct address alongside deleted ones. The answer is present, but the memory set is unsafe. We reframe agent memory as a governed object-selection problem and present MemGov, a substrate-agnostic middleware layer between existing retrieval pipelines and agent context. MemGov wraps retrieved evidence as explicit memory objects, detects must-link and cannot-link constraints, and applies deterministic constraint-aware repair, returning either a certified safe memory set or an auditable obstruction certificate rather than silently passing unsafe combinations. Evaluations on controlled governance scenarios and MemoryAgentBench-style end-to-end tasks, deliberately separating retrieval quality from governance decisions, demonstrate that while raw retrieval recovers answer-bearing evidence successfully, it simultaneously returns detected unsafe memory combinations in 47-51% of cases. The same retrieval substrates governed by MemGov reduce unsafe merges to 0% while preserving answer utility within about 2-4 percentage points, with fully auditable reasoning for all repair and refusal decisions. Additional diagnostics assess eligibility, temporal constraints, and paired answer retention.
%U https://aclanthology.org/2026.aacl-long.246/
%P 5501-5518
Markdown (Informal)
[Beyond Retrieval: Deterministic Memory Governance with Pairwise Constraints and Certified Repair for Long-Term Language Agents](https://aclanthology.org/2026.aacl-long.246/) (Christou et al., AACL-IJCNLP 2026)
ACL