Recent years have seen great successes in the development of learning algorithms in static predictive and generative tasks, where the objective is to learn a model that performs well on a single test deployment and in applications with abundant data. Comparatively less success has been achieved in designing algorithms for deployment in adaptive scenarios where the data distribution may be influenced by the choices of the algorithm itself, the algorithm needs to adaptively learn from human feedback, or the nature of the environment is rapidly changing. Yet, these are often the scenarios in which ML driven technologies will be deployed— with examples ranging from large-scale societal systems like urban traffic or social networks, to scientific experimentation, and robotics, where data is expensive, and environments are diverse and dynamic. 

Fully realizing the potential of these technologies requires both a fundamental understanding of the problems posed by low-data regimes, and dynamically evolving environments as well as the development of practical algorithms for adaptive learning. In this workshop we aim to bring together theoreticians and practitioners to discuss the pressing challenges underlying the theory, algorithms, and design of adaptive learning systems, and shed light on the promising avenues to explore to make real-world deployment of these systems a reality. Topics include but are not limited to reinforcement learning from human feedback, multi-agent reinforcement learning, foundation models for decision-making, human-robot interaction, learning from strategic environments, transfer learning in reinforcement learning, and robot learning.

Venue: TTIC Chicago, USA (July 9-11, 2024)

PROGRAM SCHEDULE