Language models generate maximally activating examples for any residual direction
Researchers develop Maximally-Activating Example Metamodels (MAEMs), language models that map a residual-stream direction from a subject model to text exemplars strongly activating it. A label-free reinforcement-learning objective maximises cosine similarity between the supplied direction and activ…
Read the full story at DEV Community — Machine Learning ↗
Timeline · 1 report
- 2026-11-07 07:01 · DEV Community — Machine Learning
Language models generate maximally activating examples for any residual direction