Fetching the paper…

Probe Pruning: Accelerating LLMs through Dynamic Pruning via Model-Probing · Around