5 papers · 1 filter
The Golden Path Hypothesis: Reusable Schedules in Diffusion Caching
Dong Wang, Wenwu Tang, Francesco Corti +3
Diffusion caching accelerates generation by replacing transformer computation with cached or predicted features at selected denoising steps. We introduce the Golden Path Hypothesis…
On the Interaction Between Model Compression and Test-Time Adaptation
Francesco Corti, Dong Wang, Young D. Kwon +2
Deep neural networks deployed in the wild must be both efficient and adaptable, requiring model compression and test-time adaptation (TTA). While both are well studied in isolation…
DLR: Zero-Inference-Cost Latent Residuals for Low-Rank Pre-Training
Dong Wang, Wenwu Tang, Yun Cheng +1
Large language models have driven recent progress in language and multimodal AI, yet pre-training them at scale is prohibitively expensive. Low-rank pre-training, which factorizes…
GRAIL: Post-hoc Compensation by Linear Reconstruction for Compressed Networks
Wenwu Tang, Dong Wang, Lothar Thiele +1
Structured deep model compression methods are hardware-friendly and substantially reduce memory and inference costs. However, under aggressive compression, the resulting accuracy d…
Cut Less, Fold More: Model Compression through the Lens of Projection Geometry
Olga Saukh, Dong Wang, Haris Šikić +2
Compressing neural networks without retraining is vital for deployment at scale. We study calibration-free compression through the lens of projection geometry: structured pruning i…