Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models

Inez Okulska; Bartosz Naskręcki; Jan Piotrowski; Tomasz Steifer

doi:10.18653/2026.findings-emnlp.1633

Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models

Inez Okulska, Bartosz Naskręcki, Jan Piotrowski, Tomasz Steifer

Correct Metadata for

Use this form to create a GitHub issue with structured data describing the correction. You will need a GitHub account. Once you create that issue, the correction will be reviewed by a staff member.

⚠️ Mobile Users: Submitting this form to create a new issue will only work with github.com, not the GitHub Mobile app.

Important: The Anthology treat PDFs as authoritative. Please use this form only to correct data that is out of line with the PDF. See our corrections guidelines if you need to change the PDF.

Title Adjust the title. Retain tags such as <fixed-case>.

Authors Adjust author names and order to match the PDF.

Abstract Correct abstract if needed. Retain XML formatting tags such as <tex-math>. You may use ... for bold, ... for italic, ... for underline, <sc>...</sc> for small-caps, <tt>...<tt> for typewriter text, <url>...</url> for URLs, <a href=...> for hyperlinks, and <par/> for paragraph breaks.

Verification against PDF Ensure that the new title/authors match the snapshot below. (If there is no snapshot or it is too small, consult the PDF.)

Authors concatenated from the text boxes above:

ALL author names match the snapshot above—including middle initials, hyphens, and accents.

Abstract

Prompt echoing is a recognized failure mode of instruct language models, in which a model instead of generating a response, mirrors the provided prompt, even though it did not receive a specific instruction to do so. Is this phenomenon a sign of the model leaking the content of its training dataset, or is it rather caused by a misaligned behavior of the internal induction/copying mechanisms? We investigate prompt echoing small language models from different families (Gemma, Llama, Qwen, SmolLM and OLMo) and show that echoing prompts are likely to have partial overlap with the training dataset but the phenomenon is primarily driven by the model’s induction heads.

Anthology ID:: 2026.findings-emnlp.1633
Volume:: Findings of the Association for Computational Linguistics: EMNLP 2026
Month:: October
Year:: 2026
Address:: Budapest, Hungary
Editors:: Hung-Yi Lee, Sunipa Dev, Jessy Li, Desmond Elliott
Venue:: Findings
SIG:: SIGDAT
Publisher:: Association for Computational Linguistics
Note:
Pages:: 34744–34750
Language:
URL:: https://aclanthology.org/2026.findings-emnlp.1633/
DOI:: 10.18653/2026.findings-emnlp.1633
Bibkey:
Cite (ACL):: Inez Okulska, Bartosz Naskręcki, Jan Piotrowski, and Tomasz Steifer. 2026. Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 34744–34750, Budapest, Hungary. Association for Computational Linguistics.
Cite (Informal):: Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models (Okulska et al., Findings 2026)
Copy Citation:
PDF:: https://aclanthology.org/2026.findings-emnlp.1633.pdf

PDF Cite Search Fix data

Export citation

BibTeX
MODS XML
Endnote
Preformatted

@inproceedings{okulska-etal-2026-mirror,
    title = "Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models",
    author = "Okulska, Inez  and
      Naskr{\k{e}}cki, Bartosz  and
      Piotrowski, Jan  and
      Steifer, Tomasz",
    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.1633/",
    doi = "10.18653/2026.findings-emnlp.1633",
    pages = "34744--34750",
    ISBN = "979-8-89176-462-0",
    abstract = "Prompt echoing is a recognized failure mode of instruct language models, in which a model instead of generating a response, mirrors the provided prompt, even though it did not receive a specific instruction to do so. Is this phenomenon a sign of the model leaking the content of its training dataset, or is it rather caused by a misaligned behavior of the internal induction/copying mechanisms? We investigate prompt echoing small language models from different families (Gemma, Llama, Qwen, SmolLM and OLMo) and show that echoing prompts are likely to have partial overlap with the training dataset but the phenomenon is primarily driven by the model{'}s induction heads."
}

Download as File

<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="okulska-etal-2026-mirror">
    <titleInfo>
        <title>Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Inez</namePart>
        <namePart type="family">Okulska</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Bartosz</namePart>
        <namePart type="family">Naskręcki</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jan</namePart>
        <namePart type="family">Piotrowski</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Tomasz</namePart>
        <namePart type="family">Steifer</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>Prompt echoing is a recognized failure mode of instruct language models, in which a model instead of generating a response, mirrors the provided prompt, even though it did not receive a specific instruction to do so. Is this phenomenon a sign of the model leaking the content of its training dataset, or is it rather caused by a misaligned behavior of the internal induction/copying mechanisms? We investigate prompt echoing small language models from different families (Gemma, Llama, Qwen, SmolLM and OLMo) and show that echoing prompts are likely to have partial overlap with the training dataset but the phenomenon is primarily driven by the model’s induction heads.</abstract>
    <identifier type="citekey">okulska-etal-2026-mirror</identifier>
    <identifier type="doi">10.18653/2026.findings-emnlp.1633</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.1633/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>34744</start>
            <end>34750</end>
        </extent>
    </part>
</mods>
</modsCollection>

Download as File

%0 Conference Proceedings
%T Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models
%A Okulska, Inez
%A Naskręcki, Bartosz
%A Piotrowski, Jan
%A Steifer, Tomasz
%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 okulska-etal-2026-mirror
%X Prompt echoing is a recognized failure mode of instruct language models, in which a model instead of generating a response, mirrors the provided prompt, even though it did not receive a specific instruction to do so. Is this phenomenon a sign of the model leaking the content of its training dataset, or is it rather caused by a misaligned behavior of the internal induction/copying mechanisms? We investigate prompt echoing small language models from different families (Gemma, Llama, Qwen, SmolLM and OLMo) and show that echoing prompts are likely to have partial overlap with the training dataset but the phenomenon is primarily driven by the model’s induction heads.
%R 10.18653/2026.findings-emnlp.1633
%U https://aclanthology.org/2026.findings-emnlp.1633/
%U https://doi.org/10.18653/2026.findings-emnlp.1633
%P 34744-34750

Download as File

Markdown (Informal)

[Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models](https://aclanthology.org/2026.findings-emnlp.1633/) (Okulska et al., Findings 2026)

Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models (Okulska et al., Findings 2026)

ACL

Inez Okulska, Bartosz Naskręcki, Jan Piotrowski, and Tomasz Steifer. 2026. Mirror, Mirror on the Wall: Prompt Echoing in Instruct Language Models. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 34744–34750, Budapest, Hungary. Association for Computational Linguistics.