Highlights from Machine Translation and Multilinguality in October 2023
Here is my monthly summary of what papers on multilinguality and machine
translation I found the most noteworthy during October 2023. There were 2,881
preprints in the computation and language category on arXiv (a new record
number), so there is a big chance that there were preprints I would like to
read that I missed.
Navigating Cultural Chasms: Exploring and Unlocking the Cultural POV of Text-To-Image Models
A preprint from Israeli Technion, Google Research, and Cambridge University
studies cultural awareness of text-to-image models. The paper has a very good
related work section that explains necessary concepts from cultural studies (I
learned a lot). Based on the theory and surveys, they collect concepts that
differ the most across cultures (e.g., holiday, teacher, wedding, protest,
king, beauty). Then, they measure