2 papers
cs.ET2026
Parameter Efficient Machine Unlearning on Hybrid Resistive Memory based Compute-in-Memory Accelerators
Ning Lin, Jichang Yang, Yangu He +17
Resistive memory compute-in-memory accelerators provide energy efficient analogue matrix vector multiplication for neural network inference, but frequent reprogramming of analogue…
cs.LG2026
DBellQuant: Breaking the Bell with Double-Bell Transformation for LLMs Post Training Binarization
Zijian Ye, Wei Huang, Yifei Yu +3
Large language models (LLMs) demonstrate remarkable performance but face substantial computational and memory challenges that limit their practical deployment. Quantization has eme…