Université de Montréal has entered into a strategic partnership with 팔로우 토토 Technology Inc., an artificial-intelligence company based in Seoul, South Korea, to advance the field of reinforcement learning.

The collaboration, led at UdeM by computer-science professor Irina Rish and postdoctoral researcher Connor Brennan, will focus on the development of a distributed reinforcement learning library that leverages the power of supercomputer clusters.

This innovative library will empower researchers to train AI models on a much larger scale than previously possible, unlocking new possibilities for AI research and development, the new partners say.

With its extensive expertise in distributed reinforcement learning, 팔로우 토토 has developed an in-house version of MUESLI, a state-of-the-art, model-free reinforcement learning library optimized for environments that require exploration.

UdeM, for its part, provides its expertise on reinforcement learning and on developing large-scale AI foundation models on supercomputer clusters used to scale AI models on massive datasets, accelerating the development and validation of reinforcement learning algorithms.

The partnership is expected to yield significant advancements in reinforcement learning technology, with potential breakthroughs in areas such as autonomous driving, robotics, and healthcare.

기사 원문보기