@inproceedings{zhao-etal-2026-neuron,
title = "Neuron-Level Emotion Control in Speech-Generative Large Audio-Language Models",
author = {Zhao, Xiutian and
Ulgen, Ismail Rasim and
Koehn, Philipp and
Schuller, Bj{\"o}rn and
Sisman, Berrak},
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {IJCNLP}-{AACL} 2026",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-aacl.97/",
pages = "1947--1961",
ISBN = "979-8-89176-444-6",
abstract = "Speech-generative large audio-language models (LALMs) can produce expressive speech, yet reliable emotion control remains elusive: conversions often miss the target affect and may degrade linguistic fidelity through refusals, hallucinations, or paraphrase. We present, to our knowledge, one of the first neuron-level studies of emotional voice conversion (EVC) in LALMs and provide intervention-based evidence that compact sets of emotion-sensitive neuron (ESN) are functionally involved in, and can be exploited for, steering expressive speech generation at inference time. ESNs are identified via success-filtered activation aggregation enforcing both emotion realization and content preservation. Across three LALMs (Qwen2.5-Omni-7B, MiniCPM-o 4.5, Kimi-Audio), ESN interventions yield emotion-specific gains that generalize to unseen speakers. Controllability depends on selector design, mask sparsity, filtering, and intervention strength. Our results provide an intervention-based framework for training-free emotion control in LALMs for EVC."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhao-etal-2026-neuron">
<titleInfo>
<title>Neuron-Level Emotion Control in Speech-Generative Large Audio-Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xiutian</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ismail</namePart>
<namePart type="given">Rasim</namePart>
<namePart type="family">Ulgen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Philipp</namePart>
<namePart type="family">Koehn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Björn</namePart>
<namePart type="family">Schuller</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Berrak</namePart>
<namePart type="family">Sisman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-444-6</identifier>
</relatedItem>
<abstract>Speech-generative large audio-language models (LALMs) can produce expressive speech, yet reliable emotion control remains elusive: conversions often miss the target affect and may degrade linguistic fidelity through refusals, hallucinations, or paraphrase. We present, to our knowledge, one of the first neuron-level studies of emotional voice conversion (EVC) in LALMs and provide intervention-based evidence that compact sets of emotion-sensitive neuron (ESN) are functionally involved in, and can be exploited for, steering expressive speech generation at inference time. ESNs are identified via success-filtered activation aggregation enforcing both emotion realization and content preservation. Across three LALMs (Qwen2.5-Omni-7B, MiniCPM-o 4.5, Kimi-Audio), ESN interventions yield emotion-specific gains that generalize to unseen speakers. Controllability depends on selector design, mask sparsity, filtering, and intervention strength. Our results provide an intervention-based framework for training-free emotion control in LALMs for EVC.</abstract>
<identifier type="citekey">zhao-etal-2026-neuron</identifier>
<location>
<url>https://aclanthology.org/2026.findings-aacl.97/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>1947</start>
<end>1961</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Neuron-Level Emotion Control in Speech-Generative Large Audio-Language Models
%A Zhao, Xiutian
%A Ulgen, Ismail Rasim
%A Koehn, Philipp
%A Schuller, Björn
%A Sisman, Berrak
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-444-6
%F zhao-etal-2026-neuron
%X Speech-generative large audio-language models (LALMs) can produce expressive speech, yet reliable emotion control remains elusive: conversions often miss the target affect and may degrade linguistic fidelity through refusals, hallucinations, or paraphrase. We present, to our knowledge, one of the first neuron-level studies of emotional voice conversion (EVC) in LALMs and provide intervention-based evidence that compact sets of emotion-sensitive neuron (ESN) are functionally involved in, and can be exploited for, steering expressive speech generation at inference time. ESNs are identified via success-filtered activation aggregation enforcing both emotion realization and content preservation. Across three LALMs (Qwen2.5-Omni-7B, MiniCPM-o 4.5, Kimi-Audio), ESN interventions yield emotion-specific gains that generalize to unseen speakers. Controllability depends on selector design, mask sparsity, filtering, and intervention strength. Our results provide an intervention-based framework for training-free emotion control in LALMs for EVC.
%U https://aclanthology.org/2026.findings-aacl.97/
%P 1947-1961
Markdown (Informal)
[Neuron-Level Emotion Control in Speech-Generative Large Audio-Language Models](https://aclanthology.org/2026.findings-aacl.97/) (Zhao et al., Findings 2026)
ACL