AI Safety Fundamentals: Alignment
A podcast by BlueDot Impact
Categorie:
83 Episodio
-
Constitutional AI Harmlessness from AI Feedback
Pubblicato: 19/7/2024 -
Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback
Pubblicato: 19/7/2024 -
Illustrating Reinforcement Learning from Human Feedback (RLHF)
Pubblicato: 19/7/2024 -
Deep Double Descent
Pubblicato: 17/6/2024 -
Chinchilla’s Wild Implications
Pubblicato: 17/6/2024 -
Eliciting Latent Knowledge
Pubblicato: 17/6/2024 -
Intro to Brain-Like-AGI Safety
Pubblicato: 17/6/2024 -
Toy Models of Superposition
Pubblicato: 17/6/2024 -
Low-Stakes Alignment
Pubblicato: 17/6/2024 -
ABS: Scanning Neural Networks for Back-Doors by Artificial Brain Stimulation
Pubblicato: 17/6/2024 -
Two-Turn Debate Doesn’t Help Humans Answer Hard Reading Comprehension Questions
Pubblicato: 17/6/2024 -
Gradient Hacking: Definitions and Examples
Pubblicato: 17/6/2024 -
Imitative Generalisation (AKA ‘Learning the Prior’)
Pubblicato: 17/6/2024 -
Discovering Latent Knowledge in Language Models Without Supervision
Pubblicato: 17/6/2024 -
Least-To-Most Prompting Enables Complex Reasoning in Large Language Models
Pubblicato: 17/6/2024 -
An Investigation of Model-Free Planning
Pubblicato: 17/6/2024 -
Empirical Findings Generalize Surprisingly Far
Pubblicato: 17/6/2024 -
Compute Trends Across Three Eras of Machine Learning
Pubblicato: 13/6/2024 -
Worst-Case Thinking in AI Alignment
Pubblicato: 29/5/2024 -
How to Get Feedback
Pubblicato: 12/5/2024
Listen to resources from the AI Safety Fundamentals: Alignment course!https://aisafetyfundamentals.com/alignment
