<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chen-etal-2026-turnflow">
    <titleInfo>
        <title>TurnFlow: Benchmarking User-State Reasoning in Multi-Turn LLM Agents via Scene-Guided User Simulation</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Zhixun</namePart>
        <namePart type="family">Chen</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ming</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Anjie</namePart>
        <namePart type="family">Liu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xiaoyuan</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Quanyu</namePart>
        <namePart type="family">Dai</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Huangxun</namePart>
        <namePart type="family">Chen</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>Large Language Models (LLMs) are increasingly deployed through conversational interfaces, yet strong single-turn performance does not imply reliability in multi-turn setting. Multi-turn assistance requires a model to maintain an evolving representation of user requirements and to reason over that representation before complying with later requests. Failures can therefore be subtle: models may fail to recognize a changed constraint or hidden conflict, or recognize it without enforcing it in the final response. Existing benchmarks primarily test fully specified prompts, long-context recall, or controlled domain environments, leaving open-domain requirement reconciliation undermeasured. We introduce TurnFlow, an interactive benchmark for evaluating whether models preserve, revise, and act on user requirements across turns. TurnFlow combines adaptive user simulation with verifiable judging, allowing conversations to unfold naturally while keeping hidden requirements and pass/fail criteria controlled. In our main evaluation, twelve current LLMs achieve only 27.5%–51.0% aggregate pass rate on 490 scenarios. The failures show that current LLMs can remain fluent and cooperative while losing the global state of the conversation, exposing a gap between conversational helpfulness and robust requirement reconciliation.</abstract>
    <identifier type="citekey">chen-etal-2026-turnflow</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.825/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>17473</start>
            <end>17497</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T TurnFlow: Benchmarking User-State Reasoning in Multi-Turn LLM Agents via Scene-Guided User Simulation
%A Chen, Zhixun
%A Li, Ming
%A Liu, Anjie
%A Zhang, Xiaoyuan
%A Dai, Quanyu
%A Chen, Huangxun
%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 chen-etal-2026-turnflow
%X Large Language Models (LLMs) are increasingly deployed through conversational interfaces, yet strong single-turn performance does not imply reliability in multi-turn setting. Multi-turn assistance requires a model to maintain an evolving representation of user requirements and to reason over that representation before complying with later requests. Failures can therefore be subtle: models may fail to recognize a changed constraint or hidden conflict, or recognize it without enforcing it in the final response. Existing benchmarks primarily test fully specified prompts, long-context recall, or controlled domain environments, leaving open-domain requirement reconciliation undermeasured. We introduce TurnFlow, an interactive benchmark for evaluating whether models preserve, revise, and act on user requirements across turns. TurnFlow combines adaptive user simulation with verifiable judging, allowing conversations to unfold naturally while keeping hidden requirements and pass/fail criteria controlled. In our main evaluation, twelve current LLMs achieve only 27.5%–51.0% aggregate pass rate on 490 scenarios. The failures show that current LLMs can remain fluent and cooperative while losing the global state of the conversation, exposing a gap between conversational helpfulness and robust requirement reconciliation.
%U https://aclanthology.org/2026.findings-emnlp.825/
%P 17473-17497

Markdown (Informal)

[TurnFlow: Benchmarking User-State Reasoning in Multi-Turn LLM Agents via Scene-Guided User Simulation](https://aclanthology.org/2026.findings-emnlp.825/) (Chen et al., Findings 2026)

• TurnFlow: Benchmarking User-State Reasoning in Multi-Turn LLM Agents via Scene-Guided User Simulation (Chen et al., Findings 2026)

ACL

• Zhixun Chen, Ming Li, Anjie Liu, Xiaoyuan Zhang, Quanyu Dai, and Huangxun Chen. 2026. TurnFlow: Benchmarking User-State Reasoning in Multi-Turn LLM Agents via Scene-Guided User Simulation. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 17473–17497, Budapest, Hungary. Association for Computational Linguistics.