{"ID":2843313,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.08092","arxiv_id":"2511.08092","title":"Pruning as Regularization: Sensitivity-Aware One-Shot Pruning in ASR","abstract":"We challenge the conventional view of neural network pruning as solely a compression technique, demonstrating that one-shot magnitude pruning serves as a powerful implicit regularizer for ASR. Using Whisper-small, we combine gradient- and Fisher-based sensitivity diagnostics with targeted, component-wise pruning. This reveals architectural asymmetries: decoder FFNs are pruning-fragile, whereas decoder self-attention and the last encoder layers contain redundancy that, when removed, improves generalization. Without fine-tuning, pruning 50% of decoder self-attention reduces WER by 2.38% absolute (20.44% relative) on LibriSpeech test-other; pruning the last four encoder layers at 50% instead yields a 1.72% absolute (14.8% relative) improvement. Gains persisted on Common Voice and TED-LIUM datasets. Beyond regularization benefits, our sensitivity-aware approach enables more aggressive one-shot compression. At 40% sparsity, where established global pruning approaches catastrophically fail, our method preserves near-baseline accuracy. This positions pruning as a first-class architectural design tool: knowing where to prune is as important as how much to prune.","short_abstract":"We challenge the conventional view of neural network pruning as solely a compression technique, demonstrating that one-shot magnitude pruning serves as a powerful implicit regularizer for ASR. Using Whisper-small, we combine gradient- and Fisher-based sensitivity diagnostics with targeted, component-wise pruning. This...","url_abs":"https://arxiv.org/abs/2511.08092","url_pdf":"https://arxiv.org/pdf/2511.08092v1","authors":"[\"Julian Irigoyen\",\"Arthur Söhler\",\"Andreas Søeborg Kirkedal\"]","published":"2025-11-11T10:45:59Z","proceeding":"eess.AS","tasks":"[\"eess.AS\",\"cs.CL\",\"cs.SD\"]","methods":"[]","has_code":false}