Unsupervised Reinforcement Learning | Nazim Bendib | AI Summit 2025 скачать в хорошем качестве

Unsupervised Reinforcement Learning | Nazim Bendib | AI Summit 2025

Join us for a deep-dive session on Unsupervised Reinforcement Learning, where we explore cutting-edge approaches to training agents without relying on manually defined reward signals. This talk will examine how unsupervised RL enables agents to learn useful behaviors and representations purely through exploration and intrinsic motivation. We'll cover foundational concepts such as skill discovery, curiosity-driven learning, and goal-conditioned policies, and discuss how these methods help scale RL to more complex and realistic environments. Whether you're a researcher, ML practitioner, or robotics enthusiast, this session will offer valuable insights into how unsupervised techniques are unlocking new possibilities in reinforcement learning. 🎙️ New to streaming or looking to level up? Check out StreamYard and get $10 discount! 😍 https://streamyard.com/pal/d/46872217...