Six Layers Less: Encoder Pruning for Whisper with Label-Free Recovery
GeoPair: Geometry-Preserving Cross-Layer Factorization for Training-Free Transformer Compression
1
Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization Problem
Neural Spectral Capacity: Measuring and Designing Architectures from Network Specification Alone
Higher-order pruning of experts in mixture-of-experts language models
1
Label-Guided Knowledge Distillation for 3D-CNNs in Action Recognition
X-AuT: Progressive Audio-Encoder Compression for Speech LLMs with Cross-Scale Distillation
1
SQS: Bayesian DNN Compression through Sparse Quantized Sub-distributions
Lightweight Vision Transformer Compression for On-Device Plant Disease Detection in Resource-Constrained Agricultural Field Conditions
Quantization-Aware Healing: a compressed, 4-bit model that outperforms its full-precision original
Making Knowledge Distillation Cheap Enough to Run at Scale
1