MIT Sloan AI Adoption: Build a Playbook That Drives Real Business ROI
Lead AI-Native Products with Microsoft's Agentic AI Program
Overview
Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Explore a comprehensive video lecture delving into the connection between linear transformers and fast weight memory systems in machine learning. Learn about the limitations of current linearized attention mechanisms and discover proposed solutions, including new update rules and kernel functions. Gain insights into the experimental results on synthetic retrieval problems, machine translation, and language modeling tasks. Understand the implications of this research for improving the efficiency and effectiveness of transformer models in deep learning applications.
Syllabus
- Intro & Overview
- Fast Weight Systems
- Distributed Storage of Symbolic Values
- Autoregressive Attention Mechanisms
- Connecting Fast Weights to Attention Mechanism
- Softmax as a Kernel Method Performer
- Linear Attention as Fast Weights
- Capacity Limitations of Linear Attention
- Synthetic Data Experimental Setup
- Improving the Update Rule
- Deterministic Parameter-Free Projection DPFP Kernel
- Experimental Results
- Conclusion & Comments
Taught by
Yannic Kilcher