..

Explainers

Explaining AI Alignment research.

Contents

• GPT-2 Teaches GPT-4: Weak-to-Strong Generalization
• How to catch an AI Liar
• Anthropic Solved Interpretability?
• Paul Christiano’s Views on AI Doom (ft. Robert Miles)
• Clarifying and prediciting AGI

GPT-2 Teaches GPT-4: Weak-to-Strong Generalization

How to catch an AI Liar

Anthropic Solved Interpretability?

Paul Christiano’s Views on AI Doom (ft. Robert Miles)

Clarifying and prediciting AGI