Attention Manifolds: Steering or Blocking Language Models by Editing Learned B-Spline Surfaces

Excerpt

arXiv:2610.00257v1 Announce Type: new Abstract: In standard transformer attention, a source token sends the same value vector to every receiver. The query determines \emph{how much} to attend but not \emph{what} to extract. This work introduces \textbf{attention manifolds}: learned 2D B-spline surfaces $S_d(q_d, k_d)$ that modulate each value dimension based on the query-key interaction. Each surface is a tensor-product cubic B-spline initialized to zero, preserving pretrained behavior. Applied