ResearchBenchmarksLLM Pruning Evaluation on PIQA and ARC-EasyFollow94.2Relative Retention (S_Com)SparseGPT17.44837.37457.377.226Jun 2, 2026Evaluation ResultsMethodMethodLinksRelative Retention (S_Com)SparseGPTCalibration=IGSP (ours...Calibration=IGSP (ours), Model Scale=LLaMA-3.1-70B, Sparsity=60% unstructured2026.0694.2SparseGPTCalibration=MetaMath,...Calibration=MetaMath, Model Scale=LLaMA-3.1-8B, Sparsity=60% unstructured2026.0681.83MagnitudeCalibration=None, Mode...Calibration=None, Model Scale=LLaMA-3.1-8B, Sparsity=60% unstructured2026.0620.4