Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore a comprehensive research talk on building intelligent agents that can continually learn, reason, and plan using neuro-symbolic concepts. Discover how this framework combines symbolic programs with modular neural networks to create compositional abstractions of the physical world, representing object properties, relations, and actions that can be flexibly reused in novel ways. Learn about the superior data efficiency and generalization capabilities demonstrated across visual reasoning tasks in 2D, 3D, motion, and video data, as well as diverse decision-making applications spanning virtual agents and real-world robotic manipulation. Gain insights into how neuro-symbolic concepts enable agents to answer queries, infer human intentions, and make long-horizon plans spanning hours to days, drawing inspiration from cognitive science theories to achieve strong performance in novel situations and for novel goals.
