@inproceedings{ren-etal-2026-modality,
title = "Modality-Specific Gating for Large Audio-Language Models",
author = "Ren, Zhenni and
Wang, Juan and
Li, Qiuchi and
Li, Lei and
Gao, Chunxiao and
Zhang, Huaping and
Zhang, Baohua",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1489/",
pages = "34051--34065",
ISBN = "979-8-89176-459-0",
abstract = "Large Audio-Language Models have made significant advances in audio-text instruction following and multimodal interaction. However, due to their text-conditioned generation architecture, current LALMs tend to be overly influenced by input text prompts. When acoustic evidence conflicts with textual prompts, incorrect or irrelevant textual cues may steer model predictions away from the audio. In this work, we systematically analyze this challenge and show that representative LALMs exhibit a persistent audio-text imbalance during decoding, where attention to audio tokens is weakened across generation steps. To address the imbalance, we propose Modality-Specific Gating (MSG), an acoustic calibration mechanism that improves audio grounding with a small number of additional parameters. MSG introduces a token-wise, query-conditioned bias to the attention logits of audio tokens before softmax, while leaving the original attention computation unchanged. With parameter-efficient fine-tuning, MSG enhances acoustic evidence utilization, leading to consistently improved performance on vocal sound classification and speech emotion recognition, as well as strong robustness under adversarial or irrelevant text prompts."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ren-etal-2026-modality">
<titleInfo>
<title>Modality-Specific Gating for Large Audio-Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zhenni</namePart>
<namePart type="family">Ren</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Juan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qiuchi</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lei</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chunxiao</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Huaping</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Baohua</namePart>
<namePart type="family">Zhang</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 Audio-Language Models have made significant advances in audio-text instruction following and multimodal interaction. However, due to their text-conditioned generation architecture, current LALMs tend to be overly influenced by input text prompts. When acoustic evidence conflicts with textual prompts, incorrect or irrelevant textual cues may steer model predictions away from the audio. In this work, we systematically analyze this challenge and show that representative LALMs exhibit a persistent audio-text imbalance during decoding, where attention to audio tokens is weakened across generation steps. To address the imbalance, we propose Modality-Specific Gating (MSG), an acoustic calibration mechanism that improves audio grounding with a small number of additional parameters. MSG introduces a token-wise, query-conditioned bias to the attention logits of audio tokens before softmax, while leaving the original attention computation unchanged. With parameter-efficient fine-tuning, MSG enhances acoustic evidence utilization, leading to consistently improved performance on vocal sound classification and speech emotion recognition, as well as strong robustness under adversarial or irrelevant text prompts.</abstract>
<identifier type="citekey">ren-etal-2026-modality</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1489/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>34051</start>
<end>34065</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Modality-Specific Gating for Large Audio-Language Models
%A Ren, Zhenni
%A Wang, Juan
%A Li, Qiuchi
%A Li, Lei
%A Gao, Chunxiao
%A Zhang, Huaping
%A Zhang, Baohua
%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 ren-etal-2026-modality
%X Large Audio-Language Models have made significant advances in audio-text instruction following and multimodal interaction. However, due to their text-conditioned generation architecture, current LALMs tend to be overly influenced by input text prompts. When acoustic evidence conflicts with textual prompts, incorrect or irrelevant textual cues may steer model predictions away from the audio. In this work, we systematically analyze this challenge and show that representative LALMs exhibit a persistent audio-text imbalance during decoding, where attention to audio tokens is weakened across generation steps. To address the imbalance, we propose Modality-Specific Gating (MSG), an acoustic calibration mechanism that improves audio grounding with a small number of additional parameters. MSG introduces a token-wise, query-conditioned bias to the attention logits of audio tokens before softmax, while leaving the original attention computation unchanged. With parameter-efficient fine-tuning, MSG enhances acoustic evidence utilization, leading to consistently improved performance on vocal sound classification and speech emotion recognition, as well as strong robustness under adversarial or irrelevant text prompts.
%U https://aclanthology.org/2026.emnlp-main.1489/
%P 34051-34065
Markdown (Informal)
[Modality-Specific Gating for Large Audio-Language Models](https://aclanthology.org/2026.emnlp-main.1489/) (Ren et al., EMNLP 2026)
ACL
- Zhenni Ren, Juan Wang, Qiuchi Li, Lei Li, Chunxiao Gao, Huaping Zhang, and Baohua Zhang. 2026. Modality-Specific Gating for Large Audio-Language Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 34051–34065, Budapest, Hungary. Association for Computational Linguistics.