Janarthanan Rajendran
ஜனார்த்தனன் ராஜேந்திரன்
Sexton Chair in Reinforcement Learning
Assistant Professor
Faculty of Computer Science
Dalhousie University
Affiliate Member
Mila - Quebec Artificial Intelligence Institute
Principal Member
AI Atlantic: The Atlantic Canada Research Consortium
E-Mail: (janarthanan.rajendran@dal.ca) | LinkedIn | Google Scholar | Dalhousie University | Mila
வணக்கம் (~ Greetings in Tamil) :)
I am currently the Sexton Chair in Reinforcement Learning and an Assistant Professor at the Faculty of Computer Science at Dalhousie University in Halifax, Nova Scotia, Canada. I am also an Affiliate Member of Mila - Quebec Artificial Intelligence Institute, Canada, and a principal member of AI Atlantic: Atlantic Canada Research Consortium.
My research interests lie primarily in reinforcement learning (RL), a branch of machine learning and artificial intelligence focused on learning from interactions and sequential decision-making. My work centers on deep reinforcement learning (deep RL), which combines deep learning, which uses artificial neural networks to learn useful representations from complex unstructured data, with RL methods. This allows RL agents to make decisions directly from high-dimensional inputs such as images, text, or rich sensor data. I am interested in both foundational research in deep RL and its interdisciplinary applications, with potential impact across sectors including business management, healthcare, transportation, energy, and manufacturing. My research so far has contributed to deep RL methods that: 1) automatically discover what to learn, 2) enable effective cooperation among multiple agents (multi-agent RL), 3) build environment models for efficient learning, 4) follow natural language instructions and assist with large language model (LLM) and dialog system training, and 5) support interdisciplinary applications, including material discovery, dynamic pricing, tuning distributed IoT systems, and optimizing ICU treatment strategies.
If you are interested in working with me at Dalhousie University, please refer to the Prospective Students page.
Prior to my current positions, I was an IVADO Postdoctoral Research Fellow at Mila - Quebec Artificial Intelligence Institute and the University of Montréal (UdeM), Canada, working with Prof. Sarath Chandar and Prof. Doina Precup. I completed my Ph.D. in Computer Science and Engineering (stream: Artificial Intelligence) at the University of Michigan (UMich), Ann Arbor, USA, working with Prof. Satinder Singh, and completed my B.Tech. and M.Tech. in Electrical Engineering (major stream: Communication Engineering, minor stream: Physics) at the Indian Institute of Technology Madras (IITM), India, working with Prof. Balaraman Ravindran and Prof. Kaushik Mitra.
Dec 2025: As part of the Canada-France partnership on AI program we are launching a project, AutoSail, funded by the French National Research Agency (ANR) and the Natural Science and Engineering Research Council of Canada (NSERC), that combines control theory and AI methods (in particular, reinforcement learning) for real-time adaptive control and navigation of sailboats. More information here.
Aug 2025: Gave a talk on 'Adapting to Local Environment Changes and Discovering Practice Rewards in Reinforcement Learning' at Cohere Labs. Video recording of the talk is available here.
Apr 2025: I will be giving a talk on 'Advancing Machine Intelligence Through Reinforcement Learning' at the 2025 Atlantic Canada AI Summit.
Apr 2025: The Atlantic Canada AI Summit 2025 is happening between May 1-3 in Halifax. I am thrilled to be part of the organizing team for this incredible event!
Jan 2025: Our work, A Generalist Hanabi Agent, Arjun V Sudhakar, Hadi Nekoei, Mathieu Reymond, Miao Liu, Janarthanan Rajendran, Sarath Chandar, has been accepted at ICLR 2025.
Nov 2024: I will be attending NeurIPS 2024 and would love to meet if you are also attending!
Sept 2024: Our work Balancing Context Length and Mixing Times for Reinforcement Learning, Matthew D Riemer, Khimya Khetarpal, Janarthanan Rajendran, Sarath Chandar, has been accepted at NeurIPS 2024.
May 2024: Join us at the inaugural Dalhousie AI Symposium on May 9th!
Apr 2024: I am happy to share that I have been awarded the 'Sexton Chair in Reinforcement Learning'.:) It is a research chair in the Faculty of Computer Science at Dalhousie University, awarded to emerging researchers having the potential to lead in their field.
Apr 2024: Our work Dynamic Incentives in Response to Dynamic Pricing, Jesse Thibodeau, Hadi Nekoei, Afaf Taïk, Janarthanan Rajendran, Golnoosh Farnadi, will be presented at the Econometric Society Interdisciplinary Frontiers (ESIF) conference on Economics and AI+ML 2024.
Jan 2024: Our work Mastering Memory Tasks with World Models, Mohammad Reza Samsami, Artem Zholus, Janarthanan Rajendran, Sarath Chandar, has been accepted for an oral presentation at ICLR 2024. :)
Jan 2024: Our work Intelligent Switching for Reset-Free RL, Darshan Patil, Janarthanan Rajendran, Glen Berseth, Sarath Chandar, has been accepted at ICLR 2024. :)
Jan 2024: Our work Learning Conditional Policies for Crystal Design Using Offline Reinforcement Learning, Prashant Govindarajan, Santiago Miret, Jarrid Rector-Brooks, Mariano Phielipp, Janarthanan Rajendran, Sarath Chandar has been accepted for the AI for Accelerated Materials Design (AI4Mat) 2023 Digital Discovery Journal Special Issue. :)
Jan 2024: Started as an Assistant Professor at the Faculty of Computer Science, Dalhousie University. :)
Nov 2023: Our work Mastering Memory Tasks with World Models, Mohammad Reza Samsami, Artem Zholus, Janarthanan Rajendran, Sarath Chandar, will be presented at the Workshop on Agent Learning in Open-Endedness (ALOE), NeurIPS 2023.
Nov 2023: Our work Learning Conditional Policies for Crystal Design Using Offline Reinforcement Learning, Prashant Govindarajan, Santiago Miret, Jarrid Rector-Brooks, Mariano Phielipp, Janarthanan Rajendran, Sarath Chandar, will be presented at the Workshop on AI for Accelerated Materials Design (AI4Mat), NeurIPS 2023.
July 2023: Our work Dealing With Non-stationarity in Decentralized Cooperative Multi-Agent Deep Reinforcement Learning via Multi-Timescale Learning, Hadi Nekoei, Akilesh Badrinaaraayanan, Amit Sinha, Mohammad Amini, Janarthanan Rajendran, Aditya Mahajan, Sarath Chandar, has been accepted at the Conference on Lifelong Learning Agents (CoLLAs) 2023.
June 2023: I will be joining the Faculty of Computer Science at Dalhousie University, Halifax, as an Assistant Professor in January 2024. :)
May 2023: Our work Conditionally Optimistic Exploration for Cooperative Deep Multi-Agent Reinforcement Learning, Xutong Zhao, Yangchen Pan, Chenjun Xiao, Sarath Chandar, Janarthanan Rajendran, has been accepted at the Conference on Uncertainty in Artificial Intelligence (UAI) 2023.
May 2023: Our work Replay Buffer with Local Forgetting for Adaptive Deep Model-Based Reinforcement Learning, Ali Rahimi-Kalahroudi, Janarthanan Rajendran, Ida Momennejad, Harm van Seijen, Sarath Chandar, has been accepted at the Conference on Lifelong Learning Agents (CoLLAs) 2023.
May 2023: Our work Towards Few-shot Coordination: Revisiting Ad-Hoc Teamplay Challenge in the Game of Hanabi, Hadi Nekoei, Xutong Zhao, Janarthanan Rajendran, Miao Liu, Sarath Chandar, has been accepted at the Conference on Lifelong Learning Agents (CoLLAs) 2023.
March 2023: Our work Conditionally Optimistic Exploration for Cooperative Deep Multi-Agent Reinforcement Learning, Xutong Zhao, Yangchen Pan, Chenjun Xiao, Sarath Chandar, Janarthanan Rajendran, is now up on arXiv.
Feb 2023: Our work Dealing With Non-stationarity in Decentralized Cooperative Multi-Agent Deep Reinforcement Learning via Multi-Timescale Learning, Hadi Nekoei, Akilesh Badrinaaraayanan, Amit Sinha, Mohammad Amini, Janarthanan Rajendran, Aditya Mahajan, Sarath Chandar, is now up on arXiv.
Nov 2022: Our work Replay Buffer with Local Forgetting for Adaptive Deep Model-Based Reinforcement Learning, Ali Rahimi-Kalahroudi, Janarthanan Rajendran, Ida Momennejad, Harm van Seijen, Sarath Chandar, has been accepted at the Workshop on Deep Reinforcement Learning at NeurIPS 2022.
Oct 2022: Our work PatchBlender: A Motion Prior for Video Transformers, Gabriele Prato, Yale Song, Janarthanan Rajendran, Devon Hjelm, Neel Joshi, Sarath Chandar, has been accepted at Workshop on Vision Transformers: Theory and Applications at NeurIPS 2022.
Aug 2022: Conference on Lifelong Learning Agents (CoLLAs) 2022!
July 2022: Our work An Introduction to Lifelong Supervised Learning, Shagun Sodhani, Mojtaba Faramarzi, Sanket Vaibhav Mehta, Pranshu Malviya, Mohamed Abdelsalam, Janarthanan Janarthanan, Sarath Chandar, is now up on arXiv.
May 2022: Our work Towards Evaluating Adaptivity of Model-Based Reinforcement Learning Methods, Yi Wan*, Ali Rahimi-Kalahroudi*, Janarthanan Rajendran, Ida Momennejad, Sarath Chandar, Harm van Seijen, has been accepted at the International Conference on Machine Learning (ICML) 2022.
Mar 2022: Our work Towards Evaluating Adaptivity of Model-Based Reinforcement Learning Methods, Yi Wan*, Ali Rahimi-Kalahroudi*, Janarthanan Rajendran, Ida Momennejad, Sarath Chandar, Harm van Seijen, has been accepted at the Workshop on Agent Learning in Open-Endedness (ALOE) at ICRL 2022.
Mar 2022: Our work Staged independent learning: Towards decentralized cooperative multi-agent Reinforcement Learning, Hadi Nekoei, Akilesh Badrinaaraayanan, Amit Sinha, Mohammad Amini, Janarthanan Rajendran, Aditya Mahajan, Sarath Chandar, has been accepted at the Workshop on Gamification and Multi-Agent Solutions at ICLR 2022.
