Published Aug 7, 2024
A recipe for frontier model post-training
Explore the transformative impact of synthetic data in AI development as Nathan Lambert delves into Reinforcement Learning from Human Feedback, emphasizing iterative training processes, data quality, and the challenging role of human data, with insights from industry leaders like Apple and Meta.
Topics covered
Popular Clips
Episode Highlights
Related Episodes
Llama 3.1 405b, Meta's AI strategy, and the new open frontier model ecosystem
Answers 383 questionsLlama 3: Scaling open LLMs to AGI
Answers 383 questionsFrontiers in synthetic data
Answers 383 questionsLlama 3.2 Vision and Molmo: Foundations for the multimodal open-source ecosystem
Answers 383 questions
Futures of the data foundry business model
Answers 383 questionsWhy reward models are still key to understanding alignment
Answers 383 questionsOpenAI's Model (behavior) Spec, RLHF transparency, and personalization questions
Answers 383 questionsA post-training approach to AI regulation with Model Specs
Answers 383 questionsA realistic path to robotic foundation models
Answers 383 questionsPhi 3 and Arctic: Outlier LMs are hints
Answers 383 questionsModel merging lessons in The Waifu Research Department
Answers 383 questionsOLMoE and the hidden simplicity in training better foundation models
Answers 383 questionsInterconnects year in review: 2023
Answers 383 questions
