research
Pruning Kept Its Coverage Promise and Shrunk the Answer Set
Calibration-Preserving Pruning treats compression as an efficiency problem after split conformal prediction fixes the reliability contract.
Summary
Calibration-Preserving Pruning treats compression as an efficiency problem after split conformal prediction fixes the reliability contract.
At 50 percent sparsity on DBpedia-14, CPP-SparseGPT reduced mean prediction-set size from 10.1 to 8.6 while accuracy moved from 0.347 to 0.366. It produced smaller sets in 13 of 15 dataset-sparsity cells, but matched controls showed generic supervised gradients explain much of the gain; the claims remain limited to reliability-sensitive classification.
Why it matters
Calibration-Preserving Pruning treats compression as an efficiency problem after split conformal prediction fixes the reliability contract.
Limits and context
- It produced smaller sets in 13 of 15 dataset-sparsity cells, but matched controls showed generic supervised gradients explain much of the gain; the claims remain limited to reliability-sensitive classification.
Key claims
Calibration-Preserving Pruning treats compression as an efficiency problem after split conformal prediction fixes the reliability contract.
Qualification: It produced smaller sets in 13 of 15 dataset-sparsity cells, but matched controls showed generic supervised gradients explain much of the gain; the claims remain limited to reliability-sensitive classification.
Evidence: source-2026-08-26-003
Sources
- arXiv preprint 2608.23744arXiv · primary research
Corrections
No corrections have been recorded for this story.