2312.06942

N/A

# link_jewx9ybwzx.pdf ## Metadata - PDFFormatVersion=1.5 - IsLinearized=false - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - CreationDate=D:20240724002503Z - Creator=LaTeX with hyperref - ModDate=D:20240724002503Z - Custom.PTEX.Fullbanner=This is pdfTeX, Version 3.141592653-2.6-1.40.25 (TeX Live 2023) kpathsea version 6.3.5 - Producer=pdfTeX-1.40.25 - Trapped=False ## Contents ### Page 1 AI CONTROL: IMPROVING SAFETY DESPITEINTENTIONAL SUBVERSIONRyan Greenblatt∗ Buck Shlegeris Kshitij Sachan Fabien RogerRedwood ResearchABSTRACTAs larg

saved by

• Clementine Yang
• Mazin K
• Arthur Pogosian

related reading

• [2312.06942] AI Control: Improving Safety Despite Intentional Subversionarxiv.org
• AI Control: Improving Safety Despite Intentional Subversion — LessWronglesswrong.com
• [2312.06942] AI Control: Improving Safety Despite Intentional Subversionar5iv.labs.arxiv.org
• AI Control: Improving Safety Despite Intentional Subversion — AI Alignment Forumalignmentforum.org
• The case for ensuring that powerful AIs are controlledblog.redwoodresearch.org
• AI Control: Improving Safety Despite Intentional Subversionarxiv.org
• The case for ensuring that powerful AIs are controlled — LessWronglesswrong.com
• Using Dangerous AI, But Safely?youtube.com
• [2502.17424] Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMsarxiv.org
• [2604.22082] Removing Sandbagging in LLMs by Training with Weak Supervisionarxiv.org
• Predicting LLM Safety Before Release by Simulating Deploymentcdn.openai.com
• Natural-emergent-misalignment-from-reward-hacking-paper.pdfassets.anthropic.com