Published Aug 15, 2024

Is ChatGPT an N-gram model on steroids?

Delve into the philosophical and technical intricacies of AI with Timothy Nguyen and Keith Duggar as they explore the distinction between describing and explaining behavior, the innovative methods for detecting overfitting in transformers, and the critical role of n-grams in AI language prediction.

Episode Highlights

Topics covered

Episode Highlights

Related Episodes

OpenAI GPT-3: Language Models are Few-Shot Learners
Answers 383 questions
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Answers 383 questions
#029 GPT-3, Prompt Engineering, Trading, AI Alignment, Intelligence
Answers 383 questions
Explainability, Reasoning, Priors and GPT-3
Answers 383 questions
#031 WE GOT ACCESS TO GPT-3! (With Gary Marcus, Walid Saba and Connor Leahy)
Answers 383 questions
NLP is not NLU and GPT-3 - Walid Saba
Answers 383 questions
#98 - Prof. LUCIANO FLORIDI - ChatGPT, Superintelligence, Ethics, Philosophy of Information
Answers 383 questions
Jürgen Schmidhuber - Neural and Non-Neural AI, Reasoning, Transformers, and LSTMs
Answers 383 questions
#039 - Lena Voita - NLP
Answers 383 questions
#032- Simon Kornblith / GoogleAI - SimCLR and Paper Haul!
Answers 383 questions
Ryan Greenblatt - Solving ARC with GPT4o
Answers 383 questions
UK Algoshambles, Neuralink, GPT-3 and Intelligence
Answers 383 questions
#51 Francois Chollet - Intelligence and Generalisation
Answers 383 questions
#73 - YASAMAN RAZEGHI & Prof. SAMEER SINGH - NLP benchmarks
Answers 383 questions
[NO MUSIC] #98 - Prof. LUCIANO FLORIDI - ChatGPT, Singularitarians, Ethics, Philosophy of Information
Answers 383 questions

Dexa/Machine Learning Street Talk (MLST)

Is ChatGPT an N-gram model on steroids?

Topics covered

Popular Clips

Overfitting Insights

Model Dynamics Exploration

Defining Operative Boundaries

AI Conversations

Template Matching in NLP

Intriguing Insights

Detecting Overfitting

Understanding Explanations

Learning Language Rules

Collaborative Agreement

Template Matching Insights

Template Matching Insights

Transformer Insights

Exploring ML Insights

Episode Highlights

Research Insights

Model Dynamics

N-gram Analysis

Related Episodes

OpenAI GPT-3: Language Models are Few-Shot Learners

Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer

#029 GPT-3, Prompt Engineering, Trading, AI Alignment, Intelligence

Explainability, Reasoning, Priors and GPT-3

#031 WE GOT ACCESS TO GPT-3! (With Gary Marcus, Walid Saba and Connor Leahy)

NLP is not NLU and GPT-3 - Walid Saba

#98 - Prof. LUCIANO FLORIDI - ChatGPT, Superintelligence, Ethics, Philosophy of Information

Jürgen Schmidhuber - Neural and Non-Neural AI, Reasoning, Transformers, and LSTMs

#039 - Lena Voita - NLP

#032- Simon Kornblith / GoogleAI - SimCLR and Paper Haul!

Ryan Greenblatt - Solving ARC with GPT4o

UK Algoshambles, Neuralink, GPT-3 and Intelligence

#51 Francois Chollet - Intelligence and Generalisation

#73 - YASAMAN RAZEGHI & Prof. SAMEER SINGH - NLP benchmarks

[NO MUSIC] #98 - Prof. LUCIANO FLORIDI - ChatGPT, Singularitarians, Ethics, Philosophy of Information

Is ChatGPT an N-gram model on steroids?

Topics covered

Popular Clips

Episode Highlights

Research InsightsThe discussion shifts to the philosophical aspects of AI, focusing on the difference between describing and explaining behavior. Timothy Nguyen and Keith Duggar explore future research directions and the importance of collaboration in advancing AI understanding.

Research Insights

Model DynamicsTimothy Nguyen provides insights into the training processes and overfitting detection in transformers, emphasizing the role of statistical measures. His novel approach to identifying overfitting without holdout sets offers a fresh perspective on model evaluation.

Model Dynamics

N-gram Analysis

Related Episodes