x

LGOICOUR

Subscribe

Message

I am Neuroscience PhD turned software developer interested in alignment and mechanistic interpretability.

19

1

5mo

LGOICOUR

Subscribe

Message

I am Neuroscience PhD turned software developer interested in alignment and mechanistic interpretability.

19

1

5mo

Where Qwen3-4B fails to report the presence of code backdoors, a linear probe succeeds

> Methods note: The code used for the experiments and related open-source repo were built with Claude. The experimental design and writeup is my own, with minimal editing and formatting amendments made with Claude. Thesis In Qwen3-4B-Instruct, there is a gap between the explicit suspicion rating it gives for backdoored...

Oct 282

Toy transformers may represent belief-state geometry optimally but not minimally

> Methods note: The code used for the experiments and related open-source repo were built with Claude. The experimental design and writeup is my own, with minimal editing and formatting amendments made with Claude. Thesis A toy transformer keeps provably predictively defunct belief state data in its residual stream. This...

Jun 2421