CFMS: Towards Explainable and Fine-Grained Chinese Multimodal Sarcasm Detection Benchmark
Junzhao ZhangHsiu-Yuan HuangChenming TangYutong Yang, Yunfang WuNational Key Laboratory for Multimedia Information Processing, Peking UniversitySchool of Computer Science, Peking UniversitySchool of Software and Microelectronics, Peking University, {zhangjunzhao, huang.hsiuyuan, tangchenming}@stu.pku.edu.cn, {yytpku, wuyf}@pku.edu.cn · 2026
Read CFMS: Towards Explainable and Fine-Grained Chinese Multimodal Sarcasm Detection Benchmark by Junzhao ZhangHsiu-Yuan HuangChenming TangYutong Yang, Yunfang WuNational Key Laboratory for Multimedia Information Processing, Peking UniversitySchool of Computer Science, Peking UniversitySchool of Software and Microelectronics, Peking University, {zhangjunzhao, huang.hsiuyuan, tangchenming}@stu.pku.edu.cn, {yytpku, wuyf}@pku.edu.cn (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-16372.txt.
Description
Zhang and colleagues built the first fine-grained Chinese dataset for spotting sarcasm across images and text. CFMS holds 2,796 image-text pairs, each annotated for whether sarcasm is present, what it targets, and how it can be explained. A parallel Chinese-English metaphor subset exposed clear limits in current models on figurative reasoning. The team also proposes a reinforcement-learning method for choosing examples that beat standard baselines, giving researchers a firmer base for sarcasm understanding.
Catalog record
| CREATOR | Junzhao ZhangHsiu-Yuan HuangChenming TangYutong Yang, Yunfang WuNational Key Laboratory for Multimedia Information Processing, Peking UniversitySchool of Computer Science, Peking UniversitySchool of Software and Microelectronics, Peking University, {zhangjunzhao, huang.hsiuyuan, tangchenming}@stu.pku.edu.cn, {yytpku, wuyf}@pku.edu.cn |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 36119 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.16372