Riemannian-Manifold Steering: Geometry-Aware Generative Autoencoders for Label-Free Steering 事件

Riemannian-Manifold Steering: Geometry-Aware Generative Autoencoders for Label-Free Steering arXiv:2605.24942v1 Announce Type: cross Abstract: Steering a language model - intervening on its internal activations to change downstream behaviour - has recently expanded beyond linear interpolation to nonlinear methods such as angular and kernelized steering, which define intervention transformations without learning an explicit geometry over paths in activation space. Freshly introduced geometry-awa