ai safety
All the articles with the tag "ai safety".
-
Proposal for tracking the effects of architecture on monitorability
via lesswrong.com
-
Astra can do a concerning amount with no chain of thought
via lesswrong.com
-
Astra and Fable still hack on simple variants of alignment evals from 2025
via lesswrong.com
-
God Help Us, Let’s Try To Learn About Mechanistic Interpretability Techniques
via astralcodexten.com
-
Planning for the end of the world, a year on
Mapping my uncertainties (April 2026)
-
Planning for the end of the world
Mapping my uncertainties (March 2025)