Abstract
We introduce update-free on-policy steering with verifiers.
Type
Publication
arXiv preprint arXiv:2603.10282
Jasper Gerigk
Computer Science PhD Student
Computer Science PhD student researching robust robot manipulation and vision-language-action models