Large language models (LLMs) are reported to be partial to certain cultures owing to the training data dominance from the English corpora. Since multilingual cultural data are often expensive to collect, existing efforts handle this by prompt engineering or culture-specific pre-training. However, they might overlook the knowledge deficiency of low-resource culture and require extensive computing resources. In this paper, we propose CultureLLM, a cost-effective solution to incorporate cultural differences into LLMs. CultureLLM adopts World Value Survey (WVS) as seed data and generates semantically equivalent training data via the proposed semantic data augmentation. Using only 50 seed samples from WVS with augmented data, we fine-tune culture-specific LLMs and one unified model (CultureLLM-One) for 9 cultures covering rich and low-resource languages. Extensive experiments on 60 culture-related datasets demonstrate that CultureLLM significantly outperforms various counterparts such as GPT-3.5 (by 8.1%) and Gemini Pro (by 9.5%) with comparable performance to GPT-4 or even better. Our human study shows that the generated samples are semantically equivalent to the original samples, providing an effective solution for LLMs augmentation. Code is released at https://github.com/Scarelette/CultureLLM.
CuMA: Aligning LLMs with Sparse Cultural Values via Demographic-Aware Mixture of Adapters
Southeast University(东南大学);ByteDance Inc.(字节跳动公司);Key Laboratory of New Generation Artificial Intelligence Technology and Its Interdisciplinary Applications (Southeast University), Ministry of Education, China(新一代人工智能技术及其交叉应用重点实验室(东南大学),中华人民共和国教育部,中国)
Do LLMs Understand Wine Descriptors Across Cultures? A Benchmark for Cultural Adaptations of Wine Reviews
Department of Computer Science, University of Copenhagen(哥本哈根大学计算机科学系);Department of Food and Resource Economics, University of Copenhagen(哥本哈根大学食品与资源经济学系);Department of Food Science, University of Copenhagen(哥本哈根大学食品科学系);Department of Computer Science, Aarhus University(阿arhus大学计算机科学系)
Does Mapo Tofu Contain Coffee? Probing LLMs for Food-related Cultural Knowledge
The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳));Shenzhen Research Institute of Big Data(深圳大数据研究院);Hebrew University of Jerusalem(耶路撒冷希伯来大学);University of Electronic Science and Technology of China(电子科技大学);University of Copenhagen(哥本哈根大学);University of Tübingen(蒂宾根大学)
CulturALL: Benchmarking Multilingual and Multicultural Competence of LLMs on Grounded Tasks
Alibaba Group(阿里巴巴集团);Beijing Language and Culture University(北京语言大学);LMU Munich(慕尼黑大学);ELLIS Institute Finland(芬兰ELLIS研究所);University of Turku(图尔库大学);IBM Research AI, UAE(阿联酋IBM人工智能研究);MBZUAI;Harbin Institute of Technology, Shenzhen(深圳哈尔滨工业大学);Southern University of Science and Technology(南方科技大学)
KAHANI: Culturally-Nuanced Visual Storytelling Tool for Non-Western Cultures
Institute for Clarity in Documentation(文档清晰度研究所);Inria Paris-Rocquencourt(法国国家信息与自动化研究所巴黎-罗康库尔中心);Rajiv Gandhi University(拉吉夫·甘地大学);Tsinghua University(清华大学);Palmer Research Laboratories(帕尔默研究实验室);Microsoft Corporation(微软公司);University of Michigan(密歇根大学);Cornell University(康奈尔大学)
Beyond Instrumental and Substitutive Paradigms: Introducing Machine Culture as an Emergent Phenomenon in Large Language Models
Institute of Neuroscience, Chinese Academy of Sciences(中国科学院神经科学研究所);University of Cambridge(剑桥大学);Chinese University of Hong Kong(香港中文大学);Institute of Public Opinion Survey, Hong Kong SAR(香港特别行政区民意调查研究所);Tsinghua University(清华大学)
Mod-Guide: An LLM-based Content Moderation Feedback System to Address Insensitive Speech toward Indigenous Ethnic and Religious Minority Communities
Department of Computer Science, University of Toronto(计算机科学系,多伦多大学);Independent University Bangladesh(孟加拉国独立大学);Indiana University(印第安纳大学);Faculty of Information, University of Toronto(信息学院,多伦多大学)
TowerVision: Understanding and Improving Multilinguality in Vision-Language Models
Instituto Superior Técnico, Universidade de Lisboa(里斯本大学技术高级学院);Instituto de Telecomunicações(电信研究所);Carnegie Mellon University(卡内基梅隆大学);Sword Health(Sword健康);TransPerfect;MICS, CentraleSupélec, Université Paris-Saclay(MICS,中央圣艾尔布里大学,巴黎萨克雷大学);ELLIS Unit Lisbon(里斯本ELLIS单位)
Estonian WinoGrande Dataset: Comparative Analysis of LLM Performance on Human and Machine Translation
TartuNLP, Institute of Computer Science(塔尔图大学计算机科学研究所TartuNLP实验室);Department of Translation Studies, Institute of Foreign Language and Cultures(塔尔图大学外语与文化研究所翻译研究系);Institute of Genomics(塔尔图大学基因组学研究所);University of Tartu(塔尔图大学)
Schema-Guided Culture-Aware Complex Event Simulation with Multi-Agent Role-Play
University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);The Private Sector Humanitarian Alliance(私营部门人道主义联盟);DEVCOM Army Research Laboratory(DEVCOM陆军研究实验室)
DRISHTIKON: A Multimodal Multilingual Benchmark for Testing Language Models' Understanding on Indian Culture
Indian Institute of Technology Patna(印度理工学院帕纳巴分校);Banasthali Vidyapeeth University(班纳萨利大学);Pandit Deendayal Energy University(德英德能源大学);Manipal University Jaipur(马哈拉施特拉邦大学贾伊普尔分校);Dwarkadas J. Sanghvi College of Engineering(德瓦尔卡斯J.桑格维工程学院)