Published Sep 16, 2020
Explainability, Reasoning, Priors and GPT-3
Drs. Tim Scarfe and Keith Duggar delve into the reasoning limits of neural networks, debating whether they represent advanced memorization or possess true reasoning capabilities, while advocating for hybrid AI models that blend symbolic reasoning with statistical methods like GPT-3 to enhance explainability and performance.
Topics covered
Popular Clips
Episode Highlights
Related Episodes
#029 GPT-3, Prompt Engineering, Trading, AI Alignment, Intelligence
Answers 383 questions
OpenAI GPT-3: Language Models are Few-Shot Learners
Answers 383 questions
Jürgen Schmidhuber - Neural and Non-Neural AI, Reasoning, Transformers, and LSTMs
Answers 383 questions
#031 WE GOT ACCESS TO GPT-3! (With Gary Marcus, Walid Saba and Connor Leahy)
Answers 383 questions#032- Simon Kornblith / GoogleAI - SimCLR and Paper Haul!
Answers 383 questions
Computation, Bayesian Model Selection, Interactive Articles
Answers 383 questions
047 Interpretable Machine Learning - Christoph Molnar
Answers 383 questionsNLP is not NLU and GPT-3 - Walid Saba
Answers 383 questions
Neel Nanda - Mechanistic Interpretability
Answers 383 questions
#73 - YASAMAN RAZEGHI & Prof. SAMEER SINGH - NLP benchmarks
Answers 383 questions
Understanding Deep Learning - Prof. SIMON PRINCE [STAFF FAVOURITE]
Answers 383 questions
Robert Lange on NN Pruning and Collective Intelligence
Answers 383 questions
#57 - Prof. Melanie Mitchell - Why AI is harder than we think
Answers 383 questions
PERPLEXITY AI - The future of search.
Answers 383 questionsKernels!
Answers 383 questions