<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ebadian-etal-2026-tampered">
    <titleInfo>
        <title>Tampered ML: Evaluating LLM Judgment of Machine Learning Research</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Soroush</namePart>
        <namePart type="family">Ebadian</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Parand</namePart>
        <namePart type="given">A</namePart>
        <namePart type="family">Alamdari</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Joe</namePart>
        <namePart type="family">Benton</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Hailey</namePart>
        <namePart type="family">Joren</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>One approach to accelerating progress on AI alignment is to automate parts of alignment research itself. Doing so requires systems that can perform reliably on hard-to-supervise research tasks, including identifying flaws in research proposals and evaluating whether a proposed experiment supports its intended claim. To evaluate this capability, we introduce **Tampered ML**, a benchmark of 1,858 item-pairs, each consisting of an experiment proposal reconstructed from a published machine learning paper and a minimally edited counterfactual that introduces a targeted methodological flaw. Models must identify the more methodologically sound proposal in both presentation orders. A human validation study on a stratified sample estimates that 80-90% of items are valid. Frontier models perform above chance but remain far from reliable (with Opus 5 achieving 77.9% accuracy). They are highly sensitive to the order of presentation and favor flawed designs that appear more specific, transparent, or conventional.</abstract>
    <identifier type="citekey">ebadian-etal-2026-tampered</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.1536/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>35058</start>
            <end>35085</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Tampered ML: Evaluating LLM Judgment of Machine Learning Research
%A Ebadian, Soroush
%A Alamdari, Parand A.
%A Benton, Joe
%A Joren, Hailey
%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 ebadian-etal-2026-tampered
%X One approach to accelerating progress on AI alignment is to automate parts of alignment research itself. Doing so requires systems that can perform reliably on hard-to-supervise research tasks, including identifying flaws in research proposals and evaluating whether a proposed experiment supports its intended claim. To evaluate this capability, we introduce **Tampered ML**, a benchmark of 1,858 item-pairs, each consisting of an experiment proposal reconstructed from a published machine learning paper and a minimally edited counterfactual that introduces a targeted methodological flaw. Models must identify the more methodologically sound proposal in both presentation orders. A human validation study on a stratified sample estimates that 80-90% of items are valid. Frontier models perform above chance but remain far from reliable (with Opus 5 achieving 77.9% accuracy). They are highly sensitive to the order of presentation and favor flawed designs that appear more specific, transparent, or conventional.
%U https://aclanthology.org/2026.emnlp-main.1536/
%P 35058-35085

Markdown (Informal)

[Tampered ML: Evaluating LLM Judgment of Machine Learning Research](https://aclanthology.org/2026.emnlp-main.1536/) (Ebadian et al., EMNLP 2026)

• Tampered ML: Evaluating LLM Judgment of Machine Learning Research (Ebadian et al., EMNLP 2026)

ACL

• Soroush Ebadian, Parand A. Alamdari, Joe Benton, and Hailey Joren. 2026. Tampered ML: Evaluating LLM Judgment of Machine Learning Research. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 35058–35085, Budapest, Hungary. Association for Computational Linguistics.