@inproceedings{ye-etal-2026-listening,
title = "Listening to the Echo: User-Reaction Aware Policy Optimization via Scalar-Verbal Hybrid Reinforcement Learning",
author = "Ye, Jing and
Zhao, Xinpei and
Xiang, Lu and
Zhang, Yaping and
Zong, Chengqing",
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.73/",
pages = "1485--1515",
ISBN = "979-8-89176-462-0",
abstract = "While current emotional support dialogue systems typically rely on expert-defined scalar rewards for alignment, these signals suffer from severe information sparsity. They cannot explain why a response failed or how to adapt to dynamic user states, often failing to capture interaction-level conversational consequences reflected in subsequent user responses. We explore whether user reactions can serve as an additional behavior-level optimization signal for emotional support alignment. Therefore, we propose Reaction Aware Policy Optimization (RAPO), a framework that optimizes over interaction consequences rather than rubric scores. RAPO treats dialogue as a reaction-driven process and utilizes simulated user responses to generate dense natural-language feedback through three core components: (1) Hindsight Dialogue Selection, which isolates pivotal turns that meaningfully alter user emotional trajectories; (2) Generative Hindsight Feedback, which transforms user reactions into contrastive ranking signals and natural-language critiques; and (3) Scalar{--}Verbal Hybrid Policy Optimization, which couples scalar reward optimization for global alignment with verbal feedback distillation for fine-grained semantic refinement. Extensive experiments on ESC and Sotopia demonstrate that RAPO significantly outperforms strong RL baselines in driving positive interaction outcomes."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ye-etal-2026-listening">
<titleInfo>
<title>Listening to the Echo: User-Reaction Aware Policy Optimization via Scalar-Verbal Hybrid Reinforcement Learning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jing</namePart>
<namePart type="family">Ye</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinpei</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Xiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yaping</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chengqing</namePart>
<namePart type="family">Zong</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>While current emotional support dialogue systems typically rely on expert-defined scalar rewards for alignment, these signals suffer from severe information sparsity. They cannot explain why a response failed or how to adapt to dynamic user states, often failing to capture interaction-level conversational consequences reflected in subsequent user responses. We explore whether user reactions can serve as an additional behavior-level optimization signal for emotional support alignment. Therefore, we propose Reaction Aware Policy Optimization (RAPO), a framework that optimizes over interaction consequences rather than rubric scores. RAPO treats dialogue as a reaction-driven process and utilizes simulated user responses to generate dense natural-language feedback through three core components: (1) Hindsight Dialogue Selection, which isolates pivotal turns that meaningfully alter user emotional trajectories; (2) Generative Hindsight Feedback, which transforms user reactions into contrastive ranking signals and natural-language critiques; and (3) Scalar–Verbal Hybrid Policy Optimization, which couples scalar reward optimization for global alignment with verbal feedback distillation for fine-grained semantic refinement. Extensive experiments on ESC and Sotopia demonstrate that RAPO significantly outperforms strong RL baselines in driving positive interaction outcomes.</abstract>
<identifier type="citekey">ye-etal-2026-listening</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.73/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>1485</start>
<end>1515</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Listening to the Echo: User-Reaction Aware Policy Optimization via Scalar-Verbal Hybrid Reinforcement Learning
%A Ye, Jing
%A Zhao, Xinpei
%A Xiang, Lu
%A Zhang, Yaping
%A Zong, Chengqing
%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 ye-etal-2026-listening
%X While current emotional support dialogue systems typically rely on expert-defined scalar rewards for alignment, these signals suffer from severe information sparsity. They cannot explain why a response failed or how to adapt to dynamic user states, often failing to capture interaction-level conversational consequences reflected in subsequent user responses. We explore whether user reactions can serve as an additional behavior-level optimization signal for emotional support alignment. Therefore, we propose Reaction Aware Policy Optimization (RAPO), a framework that optimizes over interaction consequences rather than rubric scores. RAPO treats dialogue as a reaction-driven process and utilizes simulated user responses to generate dense natural-language feedback through three core components: (1) Hindsight Dialogue Selection, which isolates pivotal turns that meaningfully alter user emotional trajectories; (2) Generative Hindsight Feedback, which transforms user reactions into contrastive ranking signals and natural-language critiques; and (3) Scalar–Verbal Hybrid Policy Optimization, which couples scalar reward optimization for global alignment with verbal feedback distillation for fine-grained semantic refinement. Extensive experiments on ESC and Sotopia demonstrate that RAPO significantly outperforms strong RL baselines in driving positive interaction outcomes.
%U https://aclanthology.org/2026.findings-emnlp.73/
%P 1485-1515
Markdown (Informal)
[Listening to the Echo: User-Reaction Aware Policy Optimization via Scalar-Verbal Hybrid Reinforcement Learning](https://aclanthology.org/2026.findings-emnlp.73/) (Ye et al., Findings 2026)
ACL