Saved in:
Bibliographic Details
Main Authors: Wang, Yihua, Jia, Qi, Xu, Cong, Chen, Feiyu, Liu, Yuhan, Zhang, Haotian, Jin, Liang, Liu, Lu, Wang, Zhichun
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.10644
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914160796762112
author Wang, Yihua
Jia, Qi
Xu, Cong
Chen, Feiyu
Liu, Yuhan
Zhang, Haotian
Jin, Liang
Liu, Lu
Wang, Zhichun
author_facet Wang, Yihua
Jia, Qi
Xu, Cong
Chen, Feiyu
Liu, Yuhan
Zhang, Haotian
Jin, Liang
Liu, Lu
Wang, Zhichun
contents Multimodal sarcasm detection is a complex task that requires distinguishing subtle complementary signals across modalities while filtering out irrelevant information. Many advanced methods rely on learning shortcuts from datasets rather than extracting intended sarcasm-related features. However, our experiments show that shortcut learning impairs the model's generalization in real-world scenarios. Furthermore, we reveal the weaknesses of current modality fusion strategies for multimodal sarcasm detection through systematic experiments, highlighting the necessity of focusing on effective modality fusion for complex emotion recognition. To address these challenges, we construct MUStARD++$^{R}$ by removing shortcut signals from MUStARD++. Then, a Multimodal Conditional Information Bottleneck (MCIB) model is introduced to enable efficient multimodal fusion for sarcasm detection. Experimental results show that the MCIB achieves the best performance without relying on shortcut learning.
format Preprint
id arxiv_https___arxiv_org_abs_2508_10644
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Conditional Information Bottleneck for Multimodal Fusion: Overcoming Shortcut Learning in Sarcasm Detection
Wang, Yihua
Jia, Qi
Xu, Cong
Chen, Feiyu
Liu, Yuhan
Zhang, Haotian
Jin, Liang
Liu, Lu
Wang, Zhichun
Machine Learning
Multimodal sarcasm detection is a complex task that requires distinguishing subtle complementary signals across modalities while filtering out irrelevant information. Many advanced methods rely on learning shortcuts from datasets rather than extracting intended sarcasm-related features. However, our experiments show that shortcut learning impairs the model's generalization in real-world scenarios. Furthermore, we reveal the weaknesses of current modality fusion strategies for multimodal sarcasm detection through systematic experiments, highlighting the necessity of focusing on effective modality fusion for complex emotion recognition. To address these challenges, we construct MUStARD++$^{R}$ by removing shortcut signals from MUStARD++. Then, a Multimodal Conditional Information Bottleneck (MCIB) model is introduced to enable efficient multimodal fusion for sarcasm detection. Experimental results show that the MCIB achieves the best performance without relying on shortcut learning.
title Conditional Information Bottleneck for Multimodal Fusion: Overcoming Shortcut Learning in Sarcasm Detection
topic Machine Learning
url https://arxiv.org/abs/2508.10644