Lindia Tjuatja

Hi! I’m Lindia, a PhD student at CMU’s Language Technologies Institute advised by Graham Neubig. I am currently interning at Apple with the Responsible AI team, and am a frequent visitor at NYU. My current research interests mainly lie in evaluation and interpretability, often with a linguistics and cognitive science bend. I also work on NLP for understudied and endangered languages.

Recently, I’ve been thinking about:

• Acquisition: While language models may often behave linguistically like humans, their inner workings are significantly different. How can we understand the process by which LMs “learn” language through targeted evaluations and methods in interpretability?
• Abstraction: How do language models represent related (syntactic) structures, both within and across (natural and artificial) languages? Can we exploit shared structures to make (multilingual) language learning more efficient?

Before CMU, I did my undergrad at UT Austin, where I studied linguistics and computer engineering. During my time there, I worked extensively with John Beavers. I’ve also been fortunate enough to have great mentorship from many others, including the amazing computational linguistics faculty there. The summer before my senior year, I was an intern at CMU working with Shruti Rijhwani.

In my (precious) free time, I enjoy painting (oil and gouache are my mediums of choice), playing the viola, hanging out with my cat, running, and reading!