✳flâneur — a map of the web's best reading

Tomek Korbak (@tomekkorbak): "I reimplemented the bliss attractor eval from Claude 4 System Card. It's fascinating how LLMs reliably fall into attractor basins of their pet obsessions, how different these attractors across LLMs, and how they say something non-trivial about LLMs' personalities. 🌀🌀🌀" | nitter.poast.org

nitter.poast.org · saved by 1 readers

A toy Inspect implementation of the Bliss Attractor eval from Claude 4 System Card Welfare Assessment - tomekkorbak/bliss-attractors

Explore this link on the map →