TY  - RPRT
TI  - HASTE: A Framework for Training-Free, Dynamic, and Steerable Compression of Pre-Trained Convolutional Neural Networks
AU  - Lukas Meiner
AU  - Jens Mehnert
AU  - Alexandru Paul Condurache
PY  - 2026
DO  - 10.1007/s42979-026-05177-0
UR  - https://arxiv.org/abs/2606.30516
ID  - 2606.30516
ER  -