Beyond Transformers - New Sequence Processing Architectures - Day 6 Morning
Center for Language & Speech Processing(CLSP), JHU via YouTube
Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Attend this comprehensive lecture and tutorial exploring evolutionary algorithms for automated design and their applications in machine learning and AI systems. Learn from experts Jan Chorowski (Pathway CTO), Marek Adamczyk (Warsaw University), and Adrian Łańcucki (NVIDIA) as they survey key ingredients of evolutionary design methods with a focus on genetic programming. Discover how evolutionary approaches handle design processes in a holistic, multi-objective way to create solutions with unique properties for programs, electronic circuits, neural networks, and antennas. Examine real-world examples of evolved solutions including approximate arithmetic circuits, neural network architectures, and image filters that demonstrate superior properties compared to conventional designs. Explore the intersection of evolutionary computing techniques with machine learning methods and understand how this emerging field of evolutionary machine learning enriches both domains. Gain insights into stochastic optimization, online optimization, mechanism design, and algorithmic game theory from researchers actively working on algorithms that predict the future and optimization under uncertainty. Benefit from the expertise of speakers who have collaborated with leading research teams including Google Brain, Microsoft Research, and prominent academic institutions.
Syllabus
[camera] Day 6 morning - JSALT 2025 - Beyond transformers - new sequence processing architectures
Taught by
Center for Language & Speech Processing(CLSP), JHU