Lune

ACL Findings2026

Lightweight Haar Wavelet Subband Pruning for LLMs

Jiang Li, Pengfei Cao, Chenxi Zhou, Tian Lan, Xiangdong Su, Kang Liu, Jun Zhao, Guanglai Gao

Abstract

Large language models (LLMs) reach state-ofthe-art performance across many NLP tasks, but their large parameter counts introduce heavy computational and memory overhead, which complicates deployment in resourceconstrained settings. Pruning is a standard compression strategy that induces sparsity to lower these costs. However, most pruning methods for LLMs depend on calibration data and expensive weight updates, which limits practical scalability. To address these limitations, we introduce Haar Wavelet Subband Pruning (HWSP), a post-training framework that requires no calibration data and no weight updates. HWSP applies a two-dimensional Haar wavelet transform to each weight matrix and decomposes it into four frequency subbands. It then assigns a uniform sparsity ratio to all subbands so that both low-and high-frequency components are retained in a balanced manner. Our theoretical analysis shows that the subband design of HWSP provides a deterministic per-subband retention guarantee, which helps mitigate the potential bias of global magnitude pruning toward dominant frequency components. Experiments on the LLaMA, OPT and Qwen model families show that HWSP achieves competitive accuracy relative to strong pruning baselines while substantially reducing pruning time. Compared with magnitude pruning, which serves as a simple calibrationfree baseline, HWSP generally achieves better downstream performance across a wide range of sparsity levels and model scales. 1 .

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • A Simple and Effective Pruning Approach for Large Language Models cites Reducing Transformer Depth on Demand with Structured Dropout
  • A Simple and Effective Pruning Approach for Large Language Models cites SparseGPT: Massive Language Models Can be Accurately Pruned in One-Shot
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on28

  1. PIQA: Reasoning about Physical Commonsense in Natural LanguageYonatan Bisk, Rowan Zellers, Ronan Le Bras, Jianfeng Gao et al.AAAI 2020 · 2,916 citations
  2. SparseGPT: Massive Language Models Can be Accurately Pruned in One-ShotElias Frantar, Dan AlistarhICML 2023 · 1,240 citations
  3. MM-Vet: Evaluating Large Multimodal Models for Integrated CapabilitiesWeihao Yu, Zhengyuan Yang, Linjie Li, Jianfeng Wang et al.ICML 2024 · 1,191 citations
  4. A Simple and Effective Pruning Approach for Large Language ModelsMingjie Sun, Zhuang Liu, Anna Bair, J. Zico KolterICLR 2024 · 794 citations
  5. Reducing Transformer Depth on Demand with Structured DropoutAngela Fan, Edouard Grave, Armand JoulinICLR 2020 · 695 citations

Related papers

  1. HBLLM: Wavelet-Enhanced High-Fidelity 1-Bit Quantization for LLMsNingning Chen, Weicai Ye, Ying JiangNeurIPS 2025 · 5 citations
  2. SwiftPrune: Hessian-Free Weight Pruning for Large Language ModelsYuhan Kang, Yang Shi, Mei Wen, Jun He et al.EMNLP Findings 2025 · 2 citations
  3. Pruning Foundation Models for High Accuracy without RetrainingPu Zhao, Fei Sun, Xuan Shen, Pinrui Yu et al.EMNLP Findings 2024 · 13 citations
  4. Learning Semi-Structured Sparsity for LLMs via Shared and Context-Aware HypernetworkLu Sun, Jun SakumaICLR 2026
  5. SparseLLM: Towards Global Pruning of Pre-trained Language ModelsGuangji Bai, Yijiang Li, Chen Ling, Kibaek Kim et al.NeurIPS 2024 · 51 citations
Dusk over the sea between two cliffs drawn in fine vertical lines