# How Sparsity Allocation Shapes Label-Free Post-Pruning Recoverability

arXiv: https://arxiv.org/abs/2605.21972 · HTML: https://mlsanity.com/paper/2605.21972/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 12 references that are in the map.
- First posted on arXiv 21 May 2026.

## Builds on (most cited first, 8 of 8)

- [The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks](https://mlsanity.com/paper/1803.03635/) (2018)
- [Learning both Weights and Connections for Efficient Neural Networks](https://mlsanity.com/paper/1506.02626/) (2015)
- [Up or Down? Adaptive Rounding for Post-Training Quantization](https://mlsanity.com/paper/2004.10568/) (2020)
- [Rigging the Lottery: Making All Tickets Winners](https://mlsanity.com/paper/1911.11134/) (2019)
- [Data-Free Quantization Through Weight Equalization and Bias Correction](https://mlsanity.com/paper/1906.04721/) (2019)
- [REPAIR: REnormalizing Permuted Activations for Interpolation Repair](https://mlsanity.com/paper/2211.08403/) (2022)
- [Post-training deep neural network pruning via layer-wise calibration](https://mlsanity.com/paper/2104.15023/) (2021)
- [Adaptive Signal Resuscitation: Channel-wise Post-Pruning Repair for Sparse Vision Networks](https://mlsanity.com/paper/2605.21426/) (2026)

## Abstract (arXiv)

Unstructured magnitude pruning at high sparsity can reduce neural network accuracy to near-random performance, while labeled retraining may be unavailable in practical deployment settings. Label-free post-pruning repair methods can partially recover collapsed sparse models, but their effectiveness depends on the sparse model left by the upstream pruning allocation. This paper studies how sparsity allocation shapes post-repair recoverability under a fixed activation-statistic repair backend. We compare ERK and LAMP allocations under the same label-free repair protocol across CIFAR-10, CIFAR-100, and Imagenette with ResNet-18, ResNet-34, and ResNet-50 at sparsities from 90% to 95.5%. The results show that allocation choice can substantially change post-repair accuracy at the same global sparsity, and that the preferred allocation varies with architecture, dataset difficulty, and sparsity level. We identify a repair-sensitive transition regime in which BatchNorm recalibration begins to fail, while activation-statistic repair still recovers nontrivial accuracy. Additional validation on ImageNet-100 and DenseNet-121 shows that the location and width of this recoverable regime depend on data scale and connectivity structure. These findings suggest that pruning allocation and post-pruning repair should be studied jointly, since the allocation determines how much activation signal remains available for label-free recovery.