AURA: Uncertainty-Routed Activation Editing for Acoustic Grounding in Speech Foundation Models
AURA is proposed, an ultra-efficient representation-editing method that freezes the pretrained model and applies sparse scale-and-shift edits to decoder cross-attention heads, supporting dynamic activation editing as a practical path for grounding AED speech models.
Natarajan Balaji Shankar, Zilai Wang, Zi-Han Wang et al.
· 0 citations