Strategy For Conditioning Generative Models - LessWrong

This post was written under the mentorship of Evan Hubinger, and assisted by discussions with Adam Jermyn and Johannes Treutlein. See also their previous posts on this project. …

x Strategy For Conditioning Generative Models — LessWrong Language Models (LLMs) Oracle AI AI Frontpage 31 Strategy For Conditioning Generative Models by james.lucassen , evhub 1st Sep 2022 AI Alignment Forum 22 min read 4 31 Ω 16 This post was written under the mentorship of Evan Hubinger, and assisted by discussions with Adam Jermyn and Johannes Treutlein. See also their previous posts on this project. Summary Conditioning Generative Models (CGM) is a strategy to accelerate alignment research using powerful semi-aligned future language models, which seems potentially promising but constraine

related reading

• Simulators — LessWronglesswrong.com
• Cyborgism — LessWronglesswrong.com
• Alignment remains a hard, unsolved problem — LessWronglesswrong.com
• Model Organisms of Misalignment: The Case for a New Pillar of Alignment Research — AI Alignment Forumalignmentforum.org
• The Case Against AI Control Research — LessWronglesswrong.com
• [2302.00805] Conditioning Predictive Models: Risks and Strategiesarxiv.org
• Conditioning, Prompts, and Fine-Tuning — LessWronglesswrong.com
• Self-Fulfilling Misalignment Data Might Be Poisoning Our AI Modelsturntrout.com
• Deep Deceptiveness — LessWronglesswrong.com
• Taking LLMs Seriously (As Language Models) — LessWronglesswrong.com
• Should We Train Against (CoT) Monitors? — LessWronglesswrong.com
• Discovering Language Model Behaviors with Model-Written Evaluations — LessWronglesswrong.com