Talk at GoodAI on Multi-Agent Perspective to AI
Published:
Posts by Collection
media
DeepMind blog on Generally Capable Agents
Published:
MIT Technology Review on Open Ended Learning
Published:
Two Minute Papers video on AI plays catch
Published:
portfolio
Abstraction aware agents
Published:
Exact encoding for RL as inference
Published:
publications
Feature selection for short text classification using Wavelet Packet Transform
Published in Conference on Computational Natural Language Learning, 2015
Anuj Mahajan, Sharmistha, Shourya Roy
Paper Link Citation
Lifted Inference Rules with Constraints
Published in Neural Information Processing Systems, 2015
Happy Mittal, Anuj Mahajan, Vibhav Gogate, Parag Singla
Paper Link Citation
Symmetry Detection and Exploitation for Function Approximation in Deep RL
Published in Conference on Autonomous Agents and MultiAgent Systems, 2017
Anuj Mahajan, Theja Tulabandhula
Paper Link Citation
Symmetry Learning for Function Approximation in Reinforcement Learning
Published in arxiv, 2017
Anuj Mahajan, Theja Tulabandhula
Paper Link Citation
Virel: A variational inference framework for reinforcement learning
Published in Neural Information Processing Systems, 2019
Anuj Mahajan*, Matthew Fellows*, Tim GJ Rudner, Shimon Whiteson
Paper Link Citation
Maven: Multi-agent variational exploration
Published in Neural Information Processing Systems, 2019
Anuj Mahajan, Tabish Rashid, Mikayel Samvelyan, Shimon Whiteson
Paper Link Citation
RODE: Learning Roles to Decompose Multi-Agent Tasks
Published in International Conference on Learning Representations, 2020
Tonghan Wang, Tarun Gupta, Anuj Mahajan, Bei Peng, Shimon Whiteson, Chongjie Zhang
Paper Link Citation
SoftDICE for Imitation Learning: Rethinking Off-policy Distribution Matching
Published in arxiv, 2021
Mingfei Sun, Anuj Mahajan, Katja Hofmann, Shimon Whiteson
Paper Link Citation
UneVEn: Universal Value Exploration for Multi-Agent Reinforcement Learning
Published in International Conference on Machine Learning, 2021
Tarun Gupta, Anuj Mahajan, Bei Peng, Wendelin Böhmer, Shimon Whiteson
Paper Link Citation
Tesseract: Tensorised Actors for Multi-Agent Reinforcement Learning
Published in International Conference on Machine Learning, 2021
Anuj Mahajan, Mikayel Samvelyan, Lei Mao, Viktor Makoviychuk, Animesh Garg, Jean Kossaifi, Shimon Whiteson, Yuke Zhu, Animashree Anandkumar
Paper Link Citation
Open-Ended Learning Leads to Generally Capable Agents
Published in DeepMind, 2021
Adam Stooke, Anuj Mahajan, Catarina Barros, Charlie Deck, Jakob Bauer, Jakub Sygnowski, Maja Trebacz, Max Jaderberg, Michael Mathieu, Nat McAleese, Nathalie Bradley-Schmieg, Nathaniel Wong, Nicolas Porcel, Roberta Raileanu, Steph Hughes-Fitt, Valentin Dalibard and Wojciech Marian Czarnecki
Paper Link Citation
Generalization in Cooperative Multi-Agent Systems
Published in arxiv, 2022
Anuj Mahajan, Mikayel Samvelyan, Tarun Gupta, Benjamin Ellis, Mingfei Sun, Tim Rocktäschel, Shimon Whiteson
Paper Link Citation
Generalization Across Observation Shifts in Reinforcement Learning
Published in arxiv, 2023
Anuj Mahajan, Amy Zhang
Paper Link Citation
talks
Virel: A variational inference framework for reinforcement learning
Published:
Talk Slides: In this talk I discuss the exact framework for encoding RL as probabilistic inference and the algorithms that arise from it.
MAVEN: Multi-Agent Variational Exploration
Published:
Talk Slides: In this talk I discuss the sub-optimality in value based methods arising from representational constraints and propose a new method MAVEN to overcome it.
A multi agent perspective to AI
Published:
Talk link: In this talk I motivate why multi-agent learning would be an important component of AI and elucidate some frameworks where it can be used in designing an AI system.
Tesseract: Tensorised Actors for Multi-agent Reinforcement Learning
Published:
Talk Slides: In this talk I discuss how tensor decompositions can be used for sample efficient Reinforcement Learning for large factored action spaces.
Open Ended Learning
Published:
In this talk I discuss how open ended learning in the XLand domain helps create generally capable AI agents, and how intent prediction for co-players in the environment can help improve generalization performance.
Multi-Agentness, Generalization and Beyond
Published:
In this talk I discuss how the problem of generalization in AI can be approached from three different directions towards creating agents that are robust to changes in environment, goals and coplayers. I present several algorithmic approaches along with analysis and applications.
Perspectives on AI Alignment
Published:
In this talk I discuss the problem of AI alignment from a generalization perspective. I cover the main issues involved in creating safe AGI for practical and scalable deployment along with several methods, analysis and applications.
teaching
Teaching Assistant
Course, IIT Delhi, 2015
TA for undergrad and graduate bridge courses, IIT Delhi for the courses:
- Machine Learning (COL774) Spring semester 2015-16
- Computer Networks (COL334) Fall semester 2015-16
Teaching Assistant
Course, University of Oxford, 2019
TA for Reinforcement Learning course floated in Hilary term for Doctoral students in Autonomous Intelligent Machines and Systems (AIMS), 2019, University of Oxford.
Tutor
Course, Hertford College, University of Oxford, 2019
Tutor for Machine learning, Trinity 2019, Hertford College, University of Oxford.