<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="javaji-etal-2026-another">
    <titleInfo>
        <title>Another Turn, Better Output? A Turn-Wise Analysis of Iterative LLM Prompting</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Shashidhar</namePart>
        <namePart type="given">Reddy</namePart>
        <namePart type="family">Javaji</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Bhavul</namePart>
        <namePart type="family">Gauri</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Zining</namePart>
        <namePart type="family">Zhu</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>Large language models are increasingly used through multi-turn refinement loops, but most evidence for iterative improvement comes from settings with informative feedback, external tools, full conversation history, or structured self-critique. Less is known about a configuration, uninformative memoryless refinement, where a model repeatedly revises only its previous output under vague instructions such as “improve it”. In this paper, we establish a framework to analyze when this uninformative memoryless refinement configuration improves LLM responses across ideation, coding, and mathematical reasoning. We evaluate seven LLMs on a 450-task pool under vague and domain-specific feedback. We score every turn and measure both task outcomes and trajectory dynamics, including semantic drift, volatility, lexical novelty, and output growth. This trajectory view reveals a consistent discovery–retention gap: models often produce better intermediate outputs before overwriting them. The effect is most severe in coding, where correctness falls from 40.7% at T1 to 2.5% at T12, while best@12 exceeds final@12 by 42.5 percentage points. Math shows model- and prompt-dependent behavior, with GPT-OSS-20B improving under elaboration but several models drifting downward. Ideation benefits from explicit steering, but vague feedback produces rhetorical inflation and feasibility loss.Our analysis stresses the importance of turn-wise, trajectory-aware evaluation of uninformative memoryless refinement and shows why verified intermediate outputs should be retained rather than automatically replaced by later revisions.</abstract>
    <identifier type="citekey">javaji-etal-2026-another</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.2337/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>53085</start>
            <end>53110</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Another Turn, Better Output? A Turn-Wise Analysis of Iterative LLM Prompting
%A Javaji, Shashidhar Reddy
%A Gauri, Bhavul
%A Zhu, Zining
%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 javaji-etal-2026-another
%X Large language models are increasingly used through multi-turn refinement loops, but most evidence for iterative improvement comes from settings with informative feedback, external tools, full conversation history, or structured self-critique. Less is known about a configuration, uninformative memoryless refinement, where a model repeatedly revises only its previous output under vague instructions such as “improve it”. In this paper, we establish a framework to analyze when this uninformative memoryless refinement configuration improves LLM responses across ideation, coding, and mathematical reasoning. We evaluate seven LLMs on a 450-task pool under vague and domain-specific feedback. We score every turn and measure both task outcomes and trajectory dynamics, including semantic drift, volatility, lexical novelty, and output growth. This trajectory view reveals a consistent discovery–retention gap: models often produce better intermediate outputs before overwriting them. The effect is most severe in coding, where correctness falls from 40.7% at T1 to 2.5% at T12, while best@12 exceeds final@12 by 42.5 percentage points. Math shows model- and prompt-dependent behavior, with GPT-OSS-20B improving under elaboration but several models drifting downward. Ideation benefits from explicit steering, but vague feedback produces rhetorical inflation and feasibility loss.Our analysis stresses the importance of turn-wise, trajectory-aware evaluation of uninformative memoryless refinement and shows why verified intermediate outputs should be retained rather than automatically replaced by later revisions.
%U https://aclanthology.org/2026.emnlp-main.2337/
%P 53085-53110

Markdown (Informal)

[Another Turn, Better Output? A Turn-Wise Analysis of Iterative LLM Prompting](https://aclanthology.org/2026.emnlp-main.2337/) (Javaji et al., EMNLP 2026)

• Another Turn, Better Output? A Turn-Wise Analysis of Iterative LLM Prompting (Javaji et al., EMNLP 2026)

ACL

• Shashidhar Reddy Javaji, Bhavul Gauri, and Zining Zhu. 2026. Another Turn, Better Output? A Turn-Wise Analysis of Iterative LLM Prompting. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 53085–53110, Budapest, Hungary. Association for Computational Linguistics.