About Me

I am a PhD student at Mila Quebec AI Institute and Concordia University in Canada with a passion for deep learning and computer vision research. My primary research interest lies in improving the efficiency of training foundation models by leveraging prior knowledge through continual learning, meta-learning, and model reuse. I am being supervised by Dr.Eugene Belilovsky. Before coming to Mila and Concordia , I worked as a research intern at KAUST in Vision CAIR group under the supervision of Dr.Mohammed Elhoseiny. I did my undergraduate at University of Moratuwa, Sri Lanka with thesis in reconstructing 3D objects with few views with Dr. Thanuja Ambegoda.

Download my CV .

Interests
  • Continual learning
  • Meta learning
  • Foundation models
  • Computer vision
Education

  • PhD in Computer Science, 2024-2028

    Concordia University

  • Masters in Computer Science, 2023-2025

    Concordia University

  • BSc in Computer science and Engineering, 2018-2023

    University of Moratuwa

News

2025

  • [2025-07-09] We released PyLO an open source optimization library to enable usage of learned optimizers in PyTorch

  • [2025-07-08] 2 papers accepted at ICML 2025 Workshops

  • [2025-05-12] 1 paper accepted at CoLLAs 2025 Main Conference (oral)

2024

  • [2024-09-29] 1 paper accepted at ECCV 2024 Workshop

  • [2024-06-01] 1 paper accepted at CVPR 2024 Main Conference

  • [2024-05-01] Fast tracked to PhD in Computer science from MS under Dr.Eugene Belilovsky

2023

  • [2023-10-03] 2 papers accepted at ICCV 2023 Main conference and workshop

  • [2023-09-01] Started my MS in Computer Science at Mila and Concordia University under Dr.Eugene Belilovsky

2022

  • [2022-10-20] 1 paper accepted at NeurIPS 2022 Workshop

Publications

PyLO: Towards Accessible Learned Optimizers in PyTorch
In ICML 2025 Workshop: Championing Open-source Development in Machine Learning.
Paul Janson, Benjamin Therien, Quentin Anthony, Xialong Huang, Abhinav Moudgil, Eugene Belilovsky
A PyTorch library that makes learned optimizers accessible to the broader ML community with CUDA-accelerated implementations with substantial speedups (5x improvement for ViT training) and integrates seamlessly with existing PyTorch workflows, enabling practical application of learned optimization to real-world large-scale tasks.
PDF Cite Code Post
PyLO: Towards Accessible Learned Optimizers in PyTorch
Beyond Cosine Decay: On the effectiveness of Infinite Learning Rate Schedule for Continual Pre-training
In Conference on Lifelong Learning Agents 2025 (oral).
Paul Janson, Vaibhav Singh, Paria Mehrbod, Adam Ibrahim, Irina Rish, Eugene Belilovsky, Benjamin Therien
We demonstrate that infinite learning rate schedules consistently outperform widely-used repeated cosine decay for continual pre-training under distribution shifts across both vision and language models, providing a more effective alternative for large-scale self-supervised learning without catastrophic forgetting.
PDF Cite Code Post
Beyond Cosine Decay: On the effectiveness of Infinite Learning Rate Schedule for Continual Pre-training
Towards motion from video diffusion models
ECCV 2024 Workshop: Foundation Models for 3D Humans (archival).
Paul Janson, Tiberiu Popa, Eugene Belilovsky
This study investigates the capabilities of video diffusion models in generating human motion from text prompts, revealing their strengths in common motions and limitations in rare or complex movements.
PDF Cite
Towards motion from video diffusion models
Overcoming Generic Knowledge Loss with Selective Parameter Update
In Computer Vision and Pattern Recognition(CVPR) 2024, Conference Track.
Wenxuan Zhang, Paul Janson, Rahaf Aljundi, Mohammed Elhoseiny
Adding knowledge to the model without destroying its generalization by finetuning small set of parameters
PDF Cite Post
Overcoming Generic Knowledge Loss with Selective Parameter Update
Continual zero-shot learning through semantically guided generative random walk
In International conference on computer vision 2023.
Wenxuan Zhang, Paul Janson, Divyansh Jha, Kai Yi, Ivan Skorodov, Mohammed Elhoseiny
Learning novel concepts, remembering previous knowledge, and adapting it to future tasks occur …
PDF Cite Code
Continual zero-shot learning through semantically guided generative random walk
Domain Aware Zero shot learning
In ICCV 2023 Workshop on out of domain generalization in computer vision.
Kai Yi, Paul Janson, Wenxuan Zhang, Mohammed Elhoseiny
Continual zero-shot learning involves learning seen classes incrementally while improving the …
PDF Cite
Domain Aware Zero shot learning
A Simple baseline that questions the use of pre-trained model in continual learning
In NeurIPS 2022 Workshop on Distribution Shifts.
Paul Janson, Wenxuan Zhang, Rahaf Aljundi, Mohammed Elhoseiny
A baseline that performs better without training in continual learning benchmarks
PDF Cite Code
A Simple baseline that questions the use of pre-trained model in continual learning

Experience

 
 
 
 
 
Concordia University
PhD Student
Concordia University
May 2024 – Present Montreal, Canada
 
 
 
 
 
Mila Quebec AI Institute
PhD Student
Mila Quebec AI Institute
May 2024 – Present Montreal, Canada
 
 
 
 
 
Mila Quebec AI Institute
Masters Student
Mila Quebec AI Institute
September 2023 – May 2024 Montreal, Canada
Fast Tracked to PhD
 
 
 
 
 
King Abdullah University of Science and Technology
Research Intern
King Abdullah University of Science and Technology
January 2021 – January 2023 Jeddah, Saudi Arabia
 
 
 
 
 
University of Moratuwa
Undergraduate Student/Researcher
University of Moratuwa
October 2018 – July 2023 Moratuwa, Sri Lanka

Contact

Please fill in the following form to leave me a message.