<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hayakawa-saggion-2026-reader">
    <titleInfo>
        <title>Reader Proficiency Shapes Layer-wise Surprisal Profiles</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Akio</namePart>
        <namePart type="family">Hayakawa</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Horacio</namePart>
        <namePart type="family">Saggion</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>Reading behaviour varies not only with linguistic input, but also with reader proficiency. In this study, we investigate whether the layer-wise relationship between surprisal from large language models (LLMs) and human gaze behaviour differs across readers with different levels of proficiency and across gaze measures. Using eye-tracking data from the MECO L2 corpus, we compare readers with high and low vocabulary proficiency on first-pass gaze duration (FPGD) and total gaze duration (TGD). We quantify the distribution of the predictive power of surprisal across model layers using Predictive Depth (PD). Across 12 tested LLMs, we find that readers with lower vocabulary proficiency tend to show deeper PD for FPGD, while this difference is smaller for TGD. Also, TGD itself shows deeper PD than FPGD in both proficiency groups. These patterns suggest that where predictive power is concentrated across LLM layers may be related to the timing and breadth of the reading processes captured by different gaze measures, and that this relationship can vary with reader proficiency. Our leave-one-out analysis further shows that the advantage of informative internal layers extends to unseen texts, although the practical improvements in prediction are limited. Overall, our results show that layer-wise LLM surprisal provides a useful perspective on variation in reading behaviour across both reader groups and gaze measures.</abstract>
    <identifier type="citekey">hayakawa-saggion-2026-reader</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.124/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>2529</start>
            <end>2541</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Reader Proficiency Shapes Layer-wise Surprisal Profiles
%A Hayakawa, Akio
%A Saggion, Horacio
%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 hayakawa-saggion-2026-reader
%X Reading behaviour varies not only with linguistic input, but also with reader proficiency. In this study, we investigate whether the layer-wise relationship between surprisal from large language models (LLMs) and human gaze behaviour differs across readers with different levels of proficiency and across gaze measures. Using eye-tracking data from the MECO L2 corpus, we compare readers with high and low vocabulary proficiency on first-pass gaze duration (FPGD) and total gaze duration (TGD). We quantify the distribution of the predictive power of surprisal across model layers using Predictive Depth (PD). Across 12 tested LLMs, we find that readers with lower vocabulary proficiency tend to show deeper PD for FPGD, while this difference is smaller for TGD. Also, TGD itself shows deeper PD than FPGD in both proficiency groups. These patterns suggest that where predictive power is concentrated across LLM layers may be related to the timing and breadth of the reading processes captured by different gaze measures, and that this relationship can vary with reader proficiency. Our leave-one-out analysis further shows that the advantage of informative internal layers extends to unseen texts, although the practical improvements in prediction are limited. Overall, our results show that layer-wise LLM surprisal provides a useful perspective on variation in reading behaviour across both reader groups and gaze measures.
%U https://aclanthology.org/2026.findings-emnlp.124/
%P 2529-2541

Markdown (Informal)

[Reader Proficiency Shapes Layer-wise Surprisal Profiles](https://aclanthology.org/2026.findings-emnlp.124/) (Hayakawa & Saggion, Findings 2026)

• Reader Proficiency Shapes Layer-wise Surprisal Profiles (Hayakawa & Saggion, Findings 2026)

ACL

• Akio Hayakawa and Horacio Saggion. 2026. Reader Proficiency Shapes Layer-wise Surprisal Profiles. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 2529–2541, Budapest, Hungary. Association for Computational Linguistics.