Beyond Size: How Gradients Shape Pruning Decisions in Large Language Models

Open in new window