Posts by Collection




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


A multi agent perspective to AI


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.

Open Ended Learning


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


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


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 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.


Course, Hertford College, University of Oxford, 2019

Tutor for Machine learning, Trinity 2019, Hertford College, University of Oxford.