@inproceedings{rana-etal-2026-oblivion,
title = "Oblivion: Self-Adaptive Agentic Memory Control through Decay-Driven Activation",
author = "Rana, Ashish and
Hung, Chia-Chien and
Sun, Qumeng and
Kunkel, Julian Martin and
Lawrence, Carolin",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2422/",
pages = "54986--55017",
ISBN = "979-8-89176-459-0",
abstract = "Human memory adapts through selective forgetting: experiences become less accessible over time but can be reactivated by reinforcement or contextual cues. In contrast, memory-augmented LLM agents rely on ``always-on'' retrieval and ``flat'' memory storage, causing high interference and latency as histories grow. We introduce Oblivion, a memory control framework that casts forgetting as decay-driven reductions in accessibility {--} not explicit deletion. Oblivion decouples memory control into read and write paths. The read path decides when to consult memory, based on agent uncertainty and memory buffer utility, avoiding redundant always-on access. The write path decides what to strengthen, by reinforcing memories contributing to forming the response. Together, this enables hierarchical memory organization that maintains persistent high-level strategies while dynamically loading details as needed. We evaluate on both static and dynamic long-horizon interaction benchmarks. Oblivion outperforms both direct and memory-augmented baselines, while reducing token cost by up to 73{\%} at 120K interaction spans. These results show that treating memory as a control problem {--} deciding when to retrieve and what to reinforce {--} is essential for sustaining long-horizon agent performance."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="rana-etal-2026-oblivion">
<titleInfo>
<title>Oblivion: Self-Adaptive Agentic Memory Control through Decay-Driven Activation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ashish</namePart>
<namePart type="family">Rana</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chia-Chien</namePart>
<namePart type="family">Hung</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qumeng</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Julian</namePart>
<namePart type="given">Martin</namePart>
<namePart type="family">Kunkel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolin</namePart>
<namePart type="family">Lawrence</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>Human memory adapts through selective forgetting: experiences become less accessible over time but can be reactivated by reinforcement or contextual cues. In contrast, memory-augmented LLM agents rely on “always-on” retrieval and “flat” memory storage, causing high interference and latency as histories grow. We introduce Oblivion, a memory control framework that casts forgetting as decay-driven reductions in accessibility – not explicit deletion. Oblivion decouples memory control into read and write paths. The read path decides when to consult memory, based on agent uncertainty and memory buffer utility, avoiding redundant always-on access. The write path decides what to strengthen, by reinforcing memories contributing to forming the response. Together, this enables hierarchical memory organization that maintains persistent high-level strategies while dynamically loading details as needed. We evaluate on both static and dynamic long-horizon interaction benchmarks. Oblivion outperforms both direct and memory-augmented baselines, while reducing token cost by up to 73% at 120K interaction spans. These results show that treating memory as a control problem – deciding when to retrieve and what to reinforce – is essential for sustaining long-horizon agent performance.</abstract>
<identifier type="citekey">rana-etal-2026-oblivion</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2422/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>54986</start>
<end>55017</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Oblivion: Self-Adaptive Agentic Memory Control through Decay-Driven Activation
%A Rana, Ashish
%A Hung, Chia-Chien
%A Sun, Qumeng
%A Kunkel, Julian Martin
%A Lawrence, Carolin
%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 rana-etal-2026-oblivion
%X Human memory adapts through selective forgetting: experiences become less accessible over time but can be reactivated by reinforcement or contextual cues. In contrast, memory-augmented LLM agents rely on “always-on” retrieval and “flat” memory storage, causing high interference and latency as histories grow. We introduce Oblivion, a memory control framework that casts forgetting as decay-driven reductions in accessibility – not explicit deletion. Oblivion decouples memory control into read and write paths. The read path decides when to consult memory, based on agent uncertainty and memory buffer utility, avoiding redundant always-on access. The write path decides what to strengthen, by reinforcing memories contributing to forming the response. Together, this enables hierarchical memory organization that maintains persistent high-level strategies while dynamically loading details as needed. We evaluate on both static and dynamic long-horizon interaction benchmarks. Oblivion outperforms both direct and memory-augmented baselines, while reducing token cost by up to 73% at 120K interaction spans. These results show that treating memory as a control problem – deciding when to retrieve and what to reinforce – is essential for sustaining long-horizon agent performance.
%U https://aclanthology.org/2026.emnlp-main.2422/
%P 54986-55017
Markdown (Informal)
[Oblivion: Self-Adaptive Agentic Memory Control through Decay-Driven Activation](https://aclanthology.org/2026.emnlp-main.2422/) (Rana et al., EMNLP 2026)
ACL