Welcome to our April 2023 monthly digest, where you can catch up with any AIhub stories you may have missed, peruse the latest news, find out about recent events, and more. This month, we learn how to address class imbalance in natural language processing, investigate personalized reward functions, and put together a list of large language model resources.
Class imbalance in training and evaluation datasets can pose a challenge for natural language processing (NLP) models, which are more heavily influenced by majority class data during training. As a result, NLP models tend to perform poorly on the minority classes, which often contain the cases that are most interesting to the downstream user. In this blogpost, Sophie Henning and Annemarie Friedrich give an overview of such class imbalance and survey methods for addressing it.
Rewards play a crucial role in reinforcement learning (RL), with the choice of reward responsible for the behaviour of an agent. However, designing reward functions is complicated. Automatically inferring a reward function is more desirable for end-users interacting with a system. Jessica Maghakian and Akanksha Saran use Interaction-Grounded Learning (IGL) to infer reward functions that capture the intent of an end-user.
Ad hoc teamwork refers to the problem of enabling an ad hoc agent to collaborate with others without prior coordination. It is representative of many real-world applications, such as the use of robots and software systems to assist humans in search and rescue. In this blogpost, Hasra Dodampegama writes about her work with Mohan Sridharan formulating ad hoc teamwork as a joint knowledge-based and data-driven reasoning and learning problem.
With the recent flurry of activity around large language models (LLMs), we’ve collected together just some of the publications on the topic. Our list includes articles, opinion pieces, videos and other resources. You can find the list, which we will update periodically, here.
In the third and final post in our series of AAAI 2023 workshop round-ups we hear from the organisers of the workshop on reinforcement learning for real-world applications, who tell us the key takeaways from their event.
The ACM SIGAI Industry Award for Excellence in Artificial Intelligence is given annually to individuals or teams who have transferred original academic research into AI applications in recent years in ways that demonstrate the power of AI techniques. Nominations are due by 31 May 2023, with the award announcement to be made on 30 June. You can find out how to nominate here.
Mila and UNESCO have joined forces on a book entitled Missing links in AI governance. Focussed on the need for better governance of AI, the book comprises 18 chapters written by academics, civil society representatives, innovators and policy makers. You can read it here.
Advancing data justice research and practice is a collaborative project which aims to augment the current thinking around data justice and to provide actionable resources that will help policymakers, practitioners, and impacted communities. As part of the project, a short series of documentaries tracks the work of the participants. The second video of the series is now available, and you can watch it here.
Published annually, the AI Index Report aims to track, collate, distil, and visualise data related to AI. The eight-chapter 2023 edition has recently been released. It investigates trends in research and development, technical performance, ethics, and public opinion. Read it here.
Another report released last month was the AI Now 2023 Landscape Report. In this publication, the authors diagnose a concentration of power in the tech industry as a pressing challenge and highlight a set of approaches that will help us confront this.
Our resources page
Forthcoming and past seminars for 2023
AI around the world focus series
UN SDGs focus series
New voices in AI series