Author

Steffen Herbold

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#machine learning Preprint Sep 2026

Gradients for Interventions and Activations for Detection: Targeted Feature Learning in Language Models

Overall, the results show that targeted feature quality depends jointly on the model signal and estimator, with detection and intervention capturing complementary properties.

Jonathan Drechsel, Steffen Herbold · 0 citations