Prashan Madumal

Prashan Madumal

Researcher and Engineer

Hello! I am a machine learning researcher in Trustworthy AI, AI safety, Explainable AI and Reinforcement Learning. As a research fellow at the University of Oxford, I explored how safe, trustworthy, fair and interpretable methods can be developed in computer vision and language models. My PhD (University of Melbourne) thesis investigated how actions, behaviours and decisions of sequential decision making agents (e.g. reinforcement learning agents) can be explained using causal models.

Interests

  • AI safety and trustworthiness
  • Explainable/Interpretable Artificial Intelligence
  • Causal Learning
  • Reinforcement Learning
  • Human-Agent Interaction

Education

  • PhD in Engineering (Artificial Intelligence)

    The University of Melbourne

  • BSc in Computer Science

    University of Colombo

Recent News

  • Dec 2020 - Our paper is accepted at AAAI 2021, preprint coming soon!
  • Oct 2020 - Honored to be a recipient of the Google PhD Fellowship, in recognition of my work in explainability of reinforcement learning agents. More details here.
  • Sep 2020 - I will be organzing the Explainable Agency in Artificial Intelligence Workshop at AAAI-21, together with Silvia Tulli, David Aha and Rosina Weber. See https://sites.google.com/view/xaiworkshop for more details.
  • May 2020 - I will be giving a Aribus research techtalk on explainable reinforcement learning.
  • Feb 2020 - Our paper “Explainable Reinforcement Learning Through a Causal Lens” will be presented as an oral talk at AAAI 2020
  • Feb 2020 - I’m volunteering for the Undergraduate Consortium at AAAI 2020 as a graduate mentor.
  • Feb 2020 - I was selected to participate in the Doctoral Consortium at AAAI 2020.

Publications

Improving Interpretability of CNN Models Using Non-Negative Concept Activation Vectors

Pre-print (2020)

Explainable Reinforcement Learning Through a Causal Lens

Proceedings of the 34th AAAI Conference on Artificial Intelligence (2020)

Explainable Agency in Reinforcement Learning Agents

Proceedings of the 34th AAAI Conference on Artificial Intelligence (2020)

Distal Explanations for Explainable Reinforcement Learning Agents

Pre-print

Designing Interactions with Intention-Aware Gaze-Enabled Artificial Agents

Human-Computer Interaction - INTERACT 2019 - 17th IFIP TC 13 International Conference (2019)

Nonverbal Communication in Human-AI Interaction: Opportunities and Challenges

Human Computer Interaction and Emerging Technologies - Adjunct Proceedings from the INTERACT 2019 Workshops (2019)

A Grounded Interaction Protocol for Explainable Artificial Intelligence

Proceedings of the 18th International Conference on Autonomous Agents and MultiAgent Systems (2019)

Explainable Agency in Intelligent Agents: Doctoral Consortium

Proceedings of the 18th International Conference on Autonomous Agents and MultiAgent Systems (2019)

Interaction design for explainable AI: workshop proposal

Proceedings of the 30th Australian Conference on Computer-Human Interaction, OzCHI (2018)

Towards a Grounded Dialog Model for Explainable Artificial Intelligence

Workshop on Socio-cognitive Systems - IJCAI (2018)

Adaptive event tree-based hybrid CEP computational model for Fog computing architecture

Proceedings of the 16th International Conference on Advances in ICT for Emerging Regions (ICTer) (2016)

Indoor navigation framework for mapping and localization of multiple robotic wheelchairs

5th International Conference on Intelligent Systems, Modelling and Simulation (2014)

Awards & Scholarships

Google PhD Fellowship

Google PhD Fellowship students are a select group recognized by Google researchers and their institutions as some of the most promising young academics in the world. The Fellowships are awarded to students who represent the future of research - link

Google PhD Travel Scholarship

AAMAS Travel Scholarship

Melbourne Research Scholarship

Available to high-achieving students undertaking graduate research at the University of Melbourne.

Hall of Fame Award

Hall of Fame award is awarded for the most outstanding engineering artifact of the year by Sysco Labs.

Microsoft Imagine Cup - World Citizenship - Local Finals 1st Runners Up

The Imagine Cup World Citizenship Competition is a global contest for the best new software to address social issues.

Motorola Inter University Innovations Championship - 1st Runners Up

Contact

Write to me - chikuwaprashan [at] gmail [dot] com