Learning to Steer, Steering to See: Unveiling the Geometry of RLVR in Large Language Models via Trainable Vectors
Vector steering finds a low-dimensional manifold behind RLVR gains and a stabilizer that preserves them.
1 stories, newest first · one card per story
Vector steering finds a low-dimensional manifold behind RLVR gains and a stabilizer that preserves them.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views