Usman Anwar

I am a PhD student in Computational and Biological Learning lab at Cambridge University, UK. I am interested in reinforcement learning, multi-agent learning and AI Safety. I am supervised by David Kruger and funded by Open Phil AI Fellowship and Vitalik Buterin Fellowship on AI Safety. Previously, I was a MS student at the Information Technology University, Pakistan, where I graduated second in my class and was awarded Graduate Student Fellowship.

Email  /  GitHub  /  Google Scholar  /  LinkedIn  /  CV

If you want to chat with me, please get in touch here.

xyz1 photo

Selected Publications

project1 image

Foundational Challenges in Assuring Alignment and Safety of Large Language Models


Usman Anwar and 41 other authors
Under Submission, 2024
arxiv / tweetprint /

This 150+ pages long agenda identifies 18 foundational challenges in assuring the alignment and safety of large language models (LLMs). These challenges are organized into three different categories: scientific understanding of LLMs, development and deployment methods, and sociotechnical challenges. Based on the identified challenges, we pose 200+, concrete research questions.

project1 image

Reward Model Ensembles Help Mitigate Overoptimization


Thomas Coste, Usman Anwar, Robert Kirk, David Krueger
Internation Conference on Learning Representations, 2024
arxiv / code /

We show that using an ensmeble of reward models can be effective in mitigating overoptimization.

project1 image

Bayesian Methods for Constraint Inference in Reinforcement Learning


Dimitris Papadimitriou, Usman Anwar, Daniel Brown
Transactions on Machine Learning Research, 2022
paper / poster /

We develop a Bayesian approach for learning constraints which provides several advantages as it can work with partial trajectories, is applicable in both stochastic and deterministic environments and due to its ability to provide a posterior distribution enables use of active learning for accurate learning of constraints.

project1 image

Inverse Constrained Reinforcement Learning


Usman Anwar*, Shehryar Malik*, Alireza Aghasi, Ali Ahmed
Internation Conference on Machine Learning, 2021
arxiv / video / code / poster / slides /

We propose a framework for learning Markovian constraints from user demonstrations in high dimensional, continuous settings. We empirically show that constraints thus learned are general and transfer well to agents with different dynamics and morphologies.


Design and source code from Leonid Keselman's website