Pruning for Large Language Models — From SparseGPT to KV-Cache Pruning31 March 2026AI Accelerator Pruning LLM SparseGPT Wanda Model Compression Sparsity KV-Cache Transformer Inference Optimization Structured Pruning Unstructured Pruning 2:4 Sparsity SliceGPT Attention Head Pruning Dynamic Sparsity