LESSWRONG
LW

Mary Phuong
309010
Message
Dialogue
Subscribe

Posts

Sorted by New

Wikitag Contributions

Comments

Sorted by
Newest
AI companies aren't really using external evaluators
Mary Phuong1y169

The latest Gemini tech report has some more info on GDM external safety testing: https://storage.googleapis.com/deepmind-media/gemini/gemini_v1_5_report.pdf 
(section 9.6, p. 71)

Reply411
No wikitag contributions to display.
48Evaluating and monitoring for AI scheming
Ω
2d
Ω
9
72Unfaithful Reasoning Can Fool Chain-of-Thought Monitoring
Ω
1mo
Ω
16
78Threat Model Literature Review
Ω
3y
Ω
4
127Clarifying AI X-risk
Ω
3y
Ω
24