Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore cutting-edge research in agentic artificial intelligence through this 19-minute conference presentation featuring groundbreaking work from MIT and Stanford researchers. Delve into two major research papers that address critical challenges in AI agent development: first, learn about DEEPSEARCH, a novel approach that overcomes reinforcement learning bottlenecks by implementing verifiable rewards through Monte Carlo Tree Search methods, developed by researchers from Stanford University, University of Tokyo, RIKEN AIP, University of Washington, UC Berkeley, and Amazon AWS. Second, examine innovative work on efficient and transferable agentic knowledge graph retrieval-augmented generation (RAG) systems enhanced through reinforcement learning techniques, presented by researchers from MIT Physics, MIT CSAIL, University of Central Florida, and IBM Research. Gain insights into how these advanced methodologies are pushing the boundaries of AI reasoning capabilities and agent intelligence, with particular focus on overcoming traditional limitations in reinforcement learning and improving knowledge graph utilization for more effective AI systems.
