Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.13024 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911384820776960 |
|---|---|
| author | Dai, Chongyuan Shen, Yaling Hu, Jinpeng Gao, Zihan Li, Jia Jiang, Yishun Wang, Yaxiong Liu, Liu Ge, Zongyuan |
| author_facet | Dai, Chongyuan Shen, Yaling Hu, Jinpeng Gao, Zihan Li, Jia Jiang, Yishun Wang, Yaxiong Liu, Liu Ge, Zongyuan |
| contents | Culture serves as a fundamental determinant of human affective processing and profoundly shapes how individuals perceive and interpret emotional stimuli. Despite this intrinsic link extant evaluations regarding cultural alignment within Large Language Models primarily prioritize declarative knowledge such as geographical facts or established societal customs. These benchmarks remain insufficient to capture the subjective interpretative variance inherent to diverse sociocultural lenses. To address this limitation, we introduce CEDAR, a multimodal benchmark constructed entirely from scenarios capturing Culturally \underline{\textsc{E}}licited \underline{\textsc{D}}istinct \underline{\textsc{A}}ffective \underline{\textsc{R}}esponses. To construct CEDAR, we implement a novel pipeline that leverages LLM-generated provisional labels to isolate instances yielding cross-cultural emotional distinctions, and subsequently derives reliable ground-truth annotations through rigorous human evaluation. The resulting benchmark comprises 10,962 instances across seven languages and 14 fine-grained emotion categories, with each language including 400 multimodal and 1,166 text-only samples. Comprehensive evaluations of 17 representative multilingual models reveal a dissociation between language consistency and cultural alignment, demonstrating that culturally grounded affective understanding remains a significant challenge for current models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_13024 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Tears or Cheers? Benchmarking LLMs via Culturally Elicited Distinct Affective Responses Dai, Chongyuan Shen, Yaling Hu, Jinpeng Gao, Zihan Li, Jia Jiang, Yishun Wang, Yaxiong Liu, Liu Ge, Zongyuan Computation and Language Culture serves as a fundamental determinant of human affective processing and profoundly shapes how individuals perceive and interpret emotional stimuli. Despite this intrinsic link extant evaluations regarding cultural alignment within Large Language Models primarily prioritize declarative knowledge such as geographical facts or established societal customs. These benchmarks remain insufficient to capture the subjective interpretative variance inherent to diverse sociocultural lenses. To address this limitation, we introduce CEDAR, a multimodal benchmark constructed entirely from scenarios capturing Culturally \underline{\textsc{E}}licited \underline{\textsc{D}}istinct \underline{\textsc{A}}ffective \underline{\textsc{R}}esponses. To construct CEDAR, we implement a novel pipeline that leverages LLM-generated provisional labels to isolate instances yielding cross-cultural emotional distinctions, and subsequently derives reliable ground-truth annotations through rigorous human evaluation. The resulting benchmark comprises 10,962 instances across seven languages and 14 fine-grained emotion categories, with each language including 400 multimodal and 1,166 text-only samples. Comprehensive evaluations of 17 representative multilingual models reveal a dissociation between language consistency and cultural alignment, demonstrating that culturally grounded affective understanding remains a significant challenge for current models. |
| title | Tears or Cheers? Benchmarking LLMs via Culturally Elicited Distinct Affective Responses |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2601.13024 |