Zheng Wei Lim

I am a postdoctoral researcher at the Institute for Logic, Language and Computation (ILLC) at the University of Amsterdam, working with Ivan Titov.

Prior to this, I obtained my PhD from the University of Melbourne, where I developed computational methods for difficult (i.e., cross-cultural or low-resource) translation. I was jointly advised by Ekaterina Vylomova, Charles Kemp and Trevor Cohn. I also worked on topics relating to multilinguality [1-6] and psycholinguistics [3-6]. In 2024, I was a student researcher at Google Research Australia.

My current research interests include:

• understanding how model architecture, training data and algorithms impose learning biases on language models, and their limitations in representing cognitively driven language phenomena
• studying how large-scale models navigate varying (sometimes conflicting) goals, and developing controllable mechanisms that drive convergent or adaptable behaviour across time, domains, languages and modalities.

Selected Publications