Concrete Problems in AI Safety
Paper
• 1606.06565 • Published
• 1
LLM-Alignment Papers
-
-
The Off-Switch Game
Paper • 1611.08219 • Published • 1 -
Learning to summarize from human feedback
Paper • 2009.01325 • Published • 4 -
Truthful AI: Developing and governing AI that does not lie
Paper • 2110.06674 • Published • 1 -
Scaling Laws for Neural Language Models
Paper • 2001.08361 • Published • 9 -
Training language models to follow instructions with human feedback
Paper • 2203.02155 • Published • 24 -
Constitutional AI: Harmlessness from AI Feedback
Paper • 2212.08073 • Published • 4 -
Discovering Language Model Behaviors with Model-Written Evaluations
Paper • 2212.09251 • Published • 1 -
Towards Bidirectional Human-AI Alignment: A Systematic Review for Clarifications, Framework, and Future Directions
Paper • 2406.09264 • Published • 2 -
Scalable AI Safety via Doubly-Efficient Debate
Paper • 2311.14125 • Published • 2