About me
I am a final year PhD candidate of the Electrical Engineering department at EPFL, and research assistant at Idiap Research Institute, in the Natural Language Understanding group under the supervision of Dr. James Henderson. My PhD research involves understanding deep attention-based models and their connection with Bayesian nonparametrics for Natural Language Processing. I received both undergraduate (2018) and masters (2020) qualifications from University of Cape Town with specializations in statistics.
I advocate for leadership by example. I believe that the values: honesty, humility and kindness are the ultimate form of optimisation – free to give and invaluable to receive. I am passionate and tenacious in my research and think that the progress direction is more important than the progress magnitude. My happiest state: rigorous mental stimulation, relaxing social interaction and regular outdoor exercise. For leisure, I enjoy spending time in nature: sunshine, sea and snowy mountains.
News
[April 2024]
Reveiwing for the journal IEEE Transactions on Pattern Analysis and Machine Intelligence
[March 2024]
Presented a seminar to the Department of Statistical Sciences at the University of Cape Town. The talk was on the evolution of NLP, the attention mechanism and my PhD Research. See the slides! here
[January 2024]
Reviewed for NAACL2024 as an emergency reviewer.
[December 2023]
Our work on Learning to Abstract with Nonparametric Variational Information Bottleneck will be presented at the Black Box NLP Workshop EMNLP2023 in Singapore!
[December 2023]
 |
Our paper Nonparametric Variational Regularisation of Pretrained Transformers is on ArXiv. |
[October 2023]
 |
Our short paper Learning to Abstract with Nonparametric Variational Information Bottleneck is accepted to findings of EMNLP 2023 in Singapore, (Paper) (Demo) (Poster) (Code) |