ABOUT ME

I am a Computer Science PhD student at the University of Waterloo, member of the Vector Institute, supervised by Professor Pascal Poupart. My research interest are broadly in alignment and post-training of LLMs with a focus on reliable and efficient machine learning algorithms.

Previously I was a Senior Research Engineer at Huawei Noah’s Ark Lab where I worked on Language Modeling, Machine Translation, Adversarial Algorithms and Model Compression. I completed my Master’s degree at McGill, working on semi-supervised learning, under the supervision of Professor Yannis Psaromiligkos and Professor Roussos Dimitrakopoulos. I was also had the privelege to work with Professor Luc Devroye and we looked at some theoretical properties of deep neural networks.

News

• 2025 - Papers on reward guided generation PARGS and FaRMA accepted at COLM and ICML respectively
• 2024 - Paper on reliable machine learning PreLoad appeared at AISTATS.
• 2023 - Awarded the Waterloo Apple PhD Fellowship for 2 years.
• 2022 - Benchmark paper on generating realistic perturbation for evaluating dialogue model robustnes NATURE accepted at NeurIPS.
• 2021 - Published our work on adversarial data augmentation for knowledge distillation Mate-KD and on Zero-Shot Knowledge Distillation at EMNLP.
• 2021 - Started PhD in Computer Science at University of Waterloo as a member of the Vector Institute.
• 2019 - Won the Best Paper Award at NAACL NeuralGen for our work on Bilingual text generation using GANs.