Ruo Yu (David) Tao

Computer Science M.Sc. @ UAlberta
Email: rtao3 "at" ualberta "dot" ca
Resume: here

I'm currently a Computing Science M.Sc. (Thesis) student at the University of Alberta, co-supervised by Adam White and Marlos C. Machado in the RLAI lab. My current research interests and current work are on agent-state learning for partially observable environments and representation learning for sequential decision making.

Previous to this, I was a First-class Honours and Dean's Honour List student studying Computer Science at McGill University. I was previously a research intern at both Mila and Microsoft Research. If you're interested in collaborating on something shoot me an email!

Publications

Novelty Search in Representational Space for Sample Efficient Exploration

Ruo Yu Tao, Vincent François-Lavet, Joelle Pineau

Oral presentation, NeurIPS 2020

Towards Solving Text-based Games by Producing Adaptive Action Spaces

Ruo Yu Tao, Marc-Alexandre Côté, Xingdi Yuan, Layla El Asri

Oral presentation, WordPlay workshop NeurIPS 2018

TextWorld: A Learning Environment for Text-based Games

Marc-Alexandre Côté, Ákos Kádár, Xingdi Yuan, Ben Kybartas, Tavian Barnes, Emery Fine, James Moore, Ruo Yu Tao, Matthew Hausknecht, Layla El Asri, Mahmoud Adada, Wendy Tay, Adam Trischler

Computer Games Workshop IJCAI 2018

Experience

National University of Singapore (NUS)

Over the summer/during the pandemic lockdown I was working on a research project under Prof. Lee Wee Sun on neural-based memory for RL and SLAM as a research assistant.

Mila (Quebec AI Institute)

I've been working on a research project in exploration in a model-based reinforcement learning setting since my final year of my undergraduate degree at McGill. The project is mentored by Vincent François-Lavet and supervised by Joelle Pineau. Our paper on this was accepted to NeurIPS 2020 for an oral presentation. Check out our paper above!

Microsoft Research Montreal

During my time at Microsoft Research, I worked on reinforcement learning and natural language problems. I was advised by Marc-Alexandre Côté on projects including:

  • Adaptive text action spaces: see the above paper!

  • TextWorld: a reinforcement learning framework for agents to learn text-based games. Also see above!

Articles etc.

  • I wrote a short piece on the dangers of reshaping in tensor manipulation libraries.

  • I was previously a contributor to an RL framework developed by the lovely people at FOR.ai. Check out the Journal of Open Source Software article here and the actual codebase itself here.

Contact

LinkedIn
Github
Come see me out people on my Twitter