Rabiul Awal

I'm a PhD student at Mila - Quebec AI Institute and University of Montreal with Aishwarya Agrawal. I was previously a visiting researcher at ServiceNow Research. I received an MSc in CS from University of Saskatchewan and BSc in CS from Noakhali Science and Technology University.

Research

I’m excited about physical AI. I work on diffusion and flow models for generation, representation learning, and control. Generative world models serve as a natural testbed for these ideas: they learn to capture the structure of physical reality through deep visual representations, and we can train policies directly in this imagined space. I’m also fascinated by scaling foundation models and reinforcement learning toward superhuman capabilities, and how these advances connect to alignment and the profound economic transformation that AI will bring.

Highlighted Work

See my Google Scholar for a full list of publications.

The Promise of RL for Autoregressive Image Editing

Saba Ahmadi*, Rabiul Awal*, Ankur Sikarwar*, Amirhossein Kazemnejad*, Ge Ya Luo, Juan Rodriguez, Sai Rajeswar, Siva Reddy, Chris Pal, Benno Krojer, Aishwarya Agrawal

NeurIPS'25

Paper Twitter

Rendering-Aware Reinforcement Learning for Vector Graphics Generation

Juan A. Rodriguez^∗, Haotian Zhang^∗, Abhay Puri, Aarash Feizi, Rishav Pramanik, Pascal Wichmann, Arnab Mondal, Mohammad Reza, Rabiul Awal + 6 others

NeurIPS'25

Paper Twitter