MOSAIC stands for Multilingual, Socio-cultural, and Agentic Intelligence in Computation. Our research at Koto MOSAIC Lab centers on Human-Centered NLP, encompassing areas such as:

• Cultural alignment of large language models (LLMs)

• Safety and Fairness in generative AI

• Agent-based frameworks for NLP applications across education, finance, healthcare, and the creative industries

• Reasoning and thinking models

• Efficient and low-resource methods for multilingual NLP

• Multimodal modeling