Kelsey Allen

Kelsey Allen is an Assistant Professor in the Department of Computer Science at the University of British Columbia (UBC) and a Senior Research Scientist at DeepMind. Her research bridges cognitive science, machine learning, and robotics, focusing on understanding and replicating human-like problem-solving, tool use, and physical reasoning. She holds a PhD from MIT (2016) under Josh Tenenbaum and a B.Sc. in Physics from UBC (2010).

Research Interests: Allen investigates computational mechanisms underlying human complex behaviors, particularly tool use and design. Her work emphasizes endowing machines with flexible problem-solving abilities. Key themes include lifelong learning, embodied cognition, and integrating symbolic and neural approaches.

Awards & Recognition

• Best Paper Award at Robotics: Science and Systems (RSS) 2018
• Oral Presentation at Cognitive Science Society 2019
• Spotlight at NeurIPS 2018 and ICLR 2019

Key Projects: Includes developing graph network simulators for rigid body dynamics, tools for physical design optimization, and studies on human tool-use learning. Her work often combines empirical experiments with machine learning models to bridge human and artificial intelligence.