@inproceedings{clark-schuler-2026-associative,
title = "Associative Memory Updates in Language Models Predict Human Reading Times",
author = "Clark, Christian and
Schuler, William",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.347/",
pages = "7267--7279",
ISBN = "979-8-89176-462-0",
abstract = "The self-attention mechanism used in transformer language models implements a form of associative, cue-based memory retrieval, which has been argued to resemble human sentence processing. However, while it is well established that surprisal estimates from transformers are predictive of human comprehension difficulty, it has been less clear how these models' internal memory operations might relate to human sentence processing. We propose to connect the two by measuring transformers' associative memory updates, which are calculated as the difference between each word{'}s predicted and actual representation within self-attention. We compute memory updates from three types of language models: (1) standard transformers; (2) CBR-RNN, a cognitively motivated model using a single self-attention head; and (3) CBR-RNN-M, a novel architecture that replaces the self-attention in CBR-RNN with a bounded-size memory cell. Associative memory updates are found to predict naturalistic human reading times over and above surprisal baselines on several psycholinguistic corpora, with evidence for more consistent effects from CBR-RNN and CBR-RNN-M than from standard transformers."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="clark-schuler-2026-associative">
<titleInfo>
<title>Associative Memory Updates in Language Models Predict Human Reading Times</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christian</namePart>
<namePart type="family">Clark</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">William</namePart>
<namePart type="family">Schuler</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>The self-attention mechanism used in transformer language models implements a form of associative, cue-based memory retrieval, which has been argued to resemble human sentence processing. However, while it is well established that surprisal estimates from transformers are predictive of human comprehension difficulty, it has been less clear how these models’ internal memory operations might relate to human sentence processing. We propose to connect the two by measuring transformers’ associative memory updates, which are calculated as the difference between each word’s predicted and actual representation within self-attention. We compute memory updates from three types of language models: (1) standard transformers; (2) CBR-RNN, a cognitively motivated model using a single self-attention head; and (3) CBR-RNN-M, a novel architecture that replaces the self-attention in CBR-RNN with a bounded-size memory cell. Associative memory updates are found to predict naturalistic human reading times over and above surprisal baselines on several psycholinguistic corpora, with evidence for more consistent effects from CBR-RNN and CBR-RNN-M than from standard transformers.</abstract>
<identifier type="citekey">clark-schuler-2026-associative</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.347/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>7267</start>
<end>7279</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Associative Memory Updates in Language Models Predict Human Reading Times
%A Clark, Christian
%A Schuler, William
%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 clark-schuler-2026-associative
%X The self-attention mechanism used in transformer language models implements a form of associative, cue-based memory retrieval, which has been argued to resemble human sentence processing. However, while it is well established that surprisal estimates from transformers are predictive of human comprehension difficulty, it has been less clear how these models’ internal memory operations might relate to human sentence processing. We propose to connect the two by measuring transformers’ associative memory updates, which are calculated as the difference between each word’s predicted and actual representation within self-attention. We compute memory updates from three types of language models: (1) standard transformers; (2) CBR-RNN, a cognitively motivated model using a single self-attention head; and (3) CBR-RNN-M, a novel architecture that replaces the self-attention in CBR-RNN with a bounded-size memory cell. Associative memory updates are found to predict naturalistic human reading times over and above surprisal baselines on several psycholinguistic corpora, with evidence for more consistent effects from CBR-RNN and CBR-RNN-M than from standard transformers.
%U https://aclanthology.org/2026.findings-emnlp.347/
%P 7267-7279
Markdown (Informal)
[Associative Memory Updates in Language Models Predict Human Reading Times](https://aclanthology.org/2026.findings-emnlp.347/) (Clark & Schuler, Findings 2026)
ACL