<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="baloni-rettinger-2026-discourse">
    <titleInfo>
        <title>Discourse-Structural Noise in LLM-Generated Explanations</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Raghvi</namePart>
        <namePart type="family">Baloni</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Achim</namePart>
        <namePart type="family">Rettinger</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 11th Workshop on Natural User-generated Text</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">JinYeong</namePart>
            <namePart type="family">Bak</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Rob</namePart>
            <namePart type="family">van der Goot</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Hyeju</namePart>
            <namePart type="family">Jang</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Weerayut</namePart>
            <namePart type="family">Buaphet</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Alan</namePart>
            <namePart type="family">Ramponi</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Wei</namePart>
            <namePart type="family">Xu</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Alan</namePart>
            <namePart type="family">Ritter</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">979-8-89176-470-5</identifier>
    </relatedItem>
    <abstract>Large language models (LLMs) are increasingly utilized to generate explanations of complex phenomena, yet existing evaluation frameworks assess quality extensively through semantic similarity to human references, a measure blind to discourse organization. We conduct a computational study of 300 ELI5 question pairs, comparing Llama-3.1 8B, Gemma-2 9B, and Mistral-7B against human-written responses using PDTB-grounded connective analysis across five relation types. All three models exhibit elaborative dominance, i.e. Elaborative connectives account for 56-66% of discourse markers versus 38% in human writing, with Causal, Contrastive, and Conditional relations systematically suppressed. This 18-28 percentage point gap persists after connective sense disambiguation (11-20 pp), ruling out measurement artifact. Across three semantic similarity metrics, PDTB connective noise explains less than 8% of shared variance (r²&lt;0.08), confirming it captures evaluation signal invisible to existing approaches. We establish discourse-structural noise as a distinct, measurable evaluation dimension for LLM-generated explanations.</abstract>
    <identifier type="citekey">baloni-rettinger-2026-discourse</identifier>
    <location>
        <url>https://aclanthology.org/2026.wnut-1.5/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>65</start>
            <end>71</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Discourse-Structural Noise in LLM-Generated Explanations
%A Baloni, Raghvi
%A Rettinger, Achim
%Y Bak, JinYeong
%Y van der Goot, Rob
%Y Jang, Hyeju
%Y Buaphet, Weerayut
%Y Ramponi, Alan
%Y Xu, Wei
%Y Ritter, Alan
%S Proceedings of the 11th Workshop on Natural User-generated Text
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-470-5
%F baloni-rettinger-2026-discourse
%X Large language models (LLMs) are increasingly utilized to generate explanations of complex phenomena, yet existing evaluation frameworks assess quality extensively through semantic similarity to human references, a measure blind to discourse organization. We conduct a computational study of 300 ELI5 question pairs, comparing Llama-3.1 8B, Gemma-2 9B, and Mistral-7B against human-written responses using PDTB-grounded connective analysis across five relation types. All three models exhibit elaborative dominance, i.e. Elaborative connectives account for 56-66% of discourse markers versus 38% in human writing, with Causal, Contrastive, and Conditional relations systematically suppressed. This 18-28 percentage point gap persists after connective sense disambiguation (11-20 pp), ruling out measurement artifact. Across three semantic similarity metrics, PDTB connective noise explains less than 8% of shared variance (r²<0.08), confirming it captures evaluation signal invisible to existing approaches. We establish discourse-structural noise as a distinct, measurable evaluation dimension for LLM-generated explanations.
%U https://aclanthology.org/2026.wnut-1.5/
%P 65-71

Markdown (Informal)

[Discourse-Structural Noise in LLM-Generated Explanations](https://aclanthology.org/2026.wnut-1.5/) (Baloni & Rettinger, WNUT 2026)

• Discourse-Structural Noise in LLM-Generated Explanations (Baloni & Rettinger, WNUT 2026)

ACL

• Raghvi Baloni and Achim Rettinger. 2026. Discourse-Structural Noise in LLM-Generated Explanations. In Proceedings of the 11th Workshop on Natural User-generated Text, pages 65–71, Budapest, Hungary. Association for Computational Linguistics.