Published Jan 15, 2024
Learning Transformer Programs with Dan Friedman - 667
Princeton's Dan Friedman delves into his research on making transformer models more interpretable by transforming them into human-readable programs, exploring philosophical and mechanistic aspects, and implementing constraints to enhance user understanding and tackle functional trade-offs.
Topics covered
Popular Clips
Episode Highlights
Related Episodes
Transformers On Large-Scale Graphs with Bayan Bruss - 641
Answers 383 questionsTransformers for Tabular Data at Capital One with Bayan Bruss - #591
Answers 383 questions
Language Modeling With State Space Models with Dan Fu - 630
Answers 383 questions
Information Extraction from Natural Document Formats with David Rosenberg - #126
Answers 383 questions
Delivering AI Systems in Highly Regulated Environments with Miriam Friedel - 653
Answers 383 questions
Towards Improved Transfer Learning with Hugo Larochelle - 631
Answers 383 questions
Deep Learning, Transformers, and the Consequences of Scale with Oriol Vinyals - #546
Answers 383 questions
Automated Machine Learning with Erez Barak - #323
Answers 383 questions
Trends in Natural Language Processing with Sebastian Ruder - TWiML Talk #216
Answers 383 questions
Deep Robotic Learning with Sergey Levine - #37
Answers 383 questions
Trends in Natural Language Processing with Nasrin Mostafazadeh - #337
Answers 383 questions
Engineering a Less Artificial Intelligence with Andreas Tolias - #379
Answers 383 questions
Language Understanding and LLMs with Christopher Manning - 686
Answers 383 questions
