Kings AI Reading Group

Welcome to the King’s AI reading group; we explore a wide variety of topics including, but not limited to:

• AI Safety and Security
• (Mechanistic) Interpretability
• Reinforcement Learning
• Out-Of-Domain Generalisation
• Meta Research
• AI Governance and Policy
• Robustness of Neural Networks
• Adversarial Attacks on Neural Networks

For an insight into the papers we have previously reviewed, please navigate to our Papers Explored and Talks and Papers sections. In this reading group, we foster a community of inquisitiveness and iterative feedback.

We are looking for a range of presentations that can cover peer-review of existing research, intriguing ideas/experiments that you would like to test the water on and overviews of recent AI Governance or Policy proposals or any other related topics/interest that you think are relevant to the Kings AI reading group. For information on joining the reading group or registering interest to present, please email Israel Mason-Williams @ israel.mason-williams@kcl.ac.uk or i.mason-williams24@imperial.ac.uk.

Currently, this reading group operates in an online format supported by Microsoft Teams; however, depending on members’ preferences, there is an appetite to support hybrid in-person attendance in the future.

We aim to publish presentation details on the website every four weeks, but if you would prefer to keep your presentation private, please contact us so that we can accommodate this request.