# Prune&Comp: Free Lunch for Layer-Pruned LLMs via Iterative Pruning with Magnitude Compensation

arXiv: https://arxiv.org/abs/2507.18212 · HTML: https://mlsanity.com/paper/2507.18212/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 30 references that are in the map.
- First posted on arXiv 24 Jul 2025.

## Builds on (most cited first, 18 of 18)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [PIQA: Reasoning about Physical Commonsense in Natural Language](https://mlsanity.com/paper/1911.11641/) (2019)
- [BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions](https://mlsanity.com/paper/1905.10044/) (2019)
- [SuperGlue: Learning Feature Matching with Graph Neural Networks](https://mlsanity.com/paper/1911.11763/) (2019)
- [Kimi k1.5: Scaling Reinforcement Learning with LLMs](https://mlsanity.com/paper/2501.12599/) (2025)
- [RACE: Large-scale ReAding Comprehension Dataset From Examinations](https://mlsanity.com/paper/1704.04683/) (2017)
- [QuaRot: Outlier-Free 4-Bit Inference in Rotated LLMs](https://mlsanity.com/paper/2404.00456/) (2024)
- [Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning](https://mlsanity.com/paper/2310.06694/) (2023)
- [SliceGPT: Compress Large Language Models by Deleting Rows and Columns](https://mlsanity.com/paper/2401.15024/) (2024)
- [ShortGPT: Layers in Large Language Models are More Redundant Than You Expect](https://mlsanity.com/paper/2403.03853/) (2024)
- [The Unreasonable Ineffectiveness of the Deeper Layers](https://mlsanity.com/paper/2403.17887/) (2024)
- [FlatQuant: Flatness Matters for LLM Quantization](https://mlsanity.com/paper/2410.09426/) (2024)
- [SLEB: Streamlining LLMs through Redundancy Verification and Elimination of Transformer Blocks](https://mlsanity.com/paper/2402.09025/) (2024)
- [LLM Pruning and Distillation in Practice: The Minitron Approach](https://mlsanity.com/paper/2408.11796/) (2024)

## Built on by (most cited first, 5 of 5)

- [What Do Visual Tokens Really Encode? Uncovering Sparsity and Redundancy in Multimodal Large Language Models](https://mlsanity.com/paper/2603.00510/) (2026)
- [Drop-Then-Recovery: How Redundant Are Vision-Language-Action Models?](https://mlsanity.com/paper/2606.27755/) (2026)
- [Rethinking Layer Redundancy: Calibration Matters More Than Search in LLM Depth Pruning](https://mlsanity.com/paper/2604.24938/) (2026)
- [Ghosted Layers: Unconstrained Activation Alignment for Recovering Layer-Pruned LLMs](https://mlsanity.com/paper/2605.15491/) (2026)
- [XMerge: Cross-Axis Selection and Reconstructive Layer Merging for LLM Depth Compression](https://mlsanity.com/paper/2609.02083/) (2026)

## Abstract (arXiv)

Layer pruning has emerged as a promising technique for compressing large language models (LLMs) while achieving acceleration proportional to the pruning ratio. In this work, we identify that removing any layer induces a significant magnitude gap in hidden states, resulting in substantial performance degradation. To address this issue, we propose Prune&Comp, a novel plug-and-play layer pruning scheme that leverages magnitude compensation to mitigate such gaps in a training-free manner. Specifically, we first estimate the magnitude gap caused by layer removal and then eliminate this gap by rescaling the remaining weights offline, with zero runtime overhead incurred. We further demonstrate the advantages of Prune&Comp through an iterative pruning strategy. When integrated with an iterative prune-and-compensate loop, Prune&Comp consistently enhances existing layer pruning metrics. For instance, when 5 layers of LLaMA-3-8B are pruned using the prevalent block influence metric, Prune&Comp nearly halves the perplexity and retains 93.19% of the original model's question-answering performance, outperforming the baseline by 4.01%.