PrototypeCommon
safety-research/assistant-axis
The Assistant Axis is a direction in activation space that captures how "Assistant-like" a model's behavior is. Models can drift away from the Assistant during conversations—sometimes toward bizarre or harmful personas. This repo contains a pipeline for generating the Assistant Axis and notebooks f…
- Owner
- safety-research (organization)
- Stars
- Stars: 182
- Forks
- Forks: 50
- Language
- Jupyter Notebook
- License
- None
- Created
- 2026
- Tier
- Prototype (100 to 249 ★) What this tier means
More repos by safety-research
All 6 repos by safety-research- safety-research/bloomStars: 1,427PythonBeta
- safety-research/persona_vectorsStars: 474PythonAlpha
- safety-research/automated-w2s-researchStars: 327PythonAlpha
- safety-research/SCONE-benchStars: 189Prototype
- safety-research/safety-toolingStars: 143PythonPrototype
Nearby in Jupyter Notebook
All Jupyter Notebook repos- smthemex/ComfyUI_TRELLISStars: 182Jupyter NotebookPrototype
- pollen-robotics/PincOpenStars: 182Jupyter NotebookPrototype
- HCIS-Lab/GaussianLSSStars: 182Jupyter NotebookPrototype
- brian-rose/ClimateModeling_coursewareStars: 182Jupyter NotebookPrototype
- benalexkeen/Introduction-to-linear-programmingStars: 182Jupyter NotebookPrototype
- leongatys/NeuralImageSynthesisStars: 182Jupyter NotebookPrototype