PrototypeCommune
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…
- Propriétaire
- safety-research (organisation)
- Étoiles
- Étoiles : 181
- Forks
- Forks : 50
- Langage
- Jupyter Notebook
- Licence
- Aucune
- Création
- 2026
- Rang
- Prototype (100 à 249 ★) Ce que veut dire ce rang
Autres dépôts de safety-research
Les 6 dépôts de safety-research- safety-research/bloomÉtoiles : 1 426PythonBêta
- safety-research/persona_vectorsÉtoiles : 474PythonAlpha
- safety-research/automated-w2s-researchÉtoiles : 327PythonAlpha
- safety-research/SCONE-benchÉtoiles : 189Prototype
- safety-research/safety-toolingÉtoiles : 142PythonPrototype
D’autres dépôts Jupyter Notebook
Tous les dépôts Jupyter Notebook- RagnarGrootKoerkamp/static-search-treeÉtoiles : 181Jupyter NotebookPrototype
- pollen-robotics/PincOpenÉtoiles : 181Jupyter NotebookPrototype
- jexp/neo4jevÉtoiles : 181Jupyter NotebookPrototype
- yoavram/CS1001.pyÉtoiles : 181Jupyter NotebookPrototype
- ricardoduarte/python-for-developersÉtoiles : 181Jupyter NotebookPrototype
- jrjohansson/numerical-python-book-codeÉtoiles : 181Jupyter NotebookPrototype