2 papers
cs.LG2026
Modality-Aware Zero-Shot Pruning and Sparse Attention for Efficient Multimodal Edge Inference
Yueyuan Sui, Payal Mohapatra, DoÄaç Eldenk +5
Edge devices increasingly run multimodal sensing pipelines that must remain accurate despite fluctuating power budgets and unpredictable sensor dropout. Existing pruning methods fa…
cs.AR2026
M2XFP: A Metadata-Augmented Microscaling Data Format for Efficient Low-bit Quantization
Weiming Hu, Zihan Zhang, Haoyan Zhang +8
Existing low-bit Microscaling (MX) formats, such as MXFP4, often suffer from substantial accuracy degradation due to the use of a shared scaling factor with the Power-of-Two format…