I am a research intern at Mila, Quebec AI Institute. I am affiliated with MAIR lab, supervised by professor Aishwarya Agrawal. I completed my masters at University of Saskatchewan. During that period, I also worked at SocialAI.Studio at Singapore University of Technology and Design supervised by professor Roy Ka-Wei Lee. Prior to moving to Canada, I was a software engineer at a Nordic startup with 2 years of experience in Python, Pytorch, NoSQL, microservices, and AWS. I completed my bachelor’s in Computer Science at Noakhali Science and Technology Universiy.
My research interests are in language+vision AI, compositionality and representation learning and Social NLP. I’ve published 3 papers in conference venues and lead an NLP research reading group in Dhaka.
My current research goal is studying compositionality and reasoning in deep language+vision and out-of-distribution generalization in neural networks.
The goal of this project is to systematically evaluate the generalization of pre-trained vision-language models (VLMs) in out-of-distribution (OOD) settings. I aim to achieve this through (1) investigating the reasoning skills of state-of-the-art (SOTA) pre-trained vision-language Transformers and (2) systematic probing of VLMs for compositionality.
My early research focuses on deep learning applications for tackling two social value tasks: online misinformation and hate speech in multiple languages.
I occasionally write on how-to-cs-grad, coding, ai research, and systematic failure in Bangladesh.
Please send me an email if you are interested in collaboration or need help with grad apps and how-to-cs-undergrad.
rabiul.awal @ mila dot quebec
awalrabiul6 @ gmail dot com
updated october 2022