mlsanityData to 14 Oct 2026

PLUM: Improving Inference Efficiency By Leveraging Repetition-Sparsity Trade-Off

arXiv 2312.01581 ↗First posted 4 Dec 2023cs.LG cs.AI cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Efficient inference of Deep Neural Networks (DNNs) on resource-constrained edge devices is essential. Quantization and sparsity are key techniques that translate to repetition and sparsity within tensors at the hardware-software interface. This paper introduces the concept of repetition-sparsity trade-off that helps explain computational efficiency during inference. We propose PLUM, a unified co-design framework that integrates DNN inference systems and quantization (forward and backward pass) to leverage the repetition-sparsity trade-off to improve inference efficiency. Our results demonstrate that PLUM's quantization method is more accurate than binary quantization with the same number of non-zero weights. Detailed analysis indicates that signed binarization generates a smaller distribution of effectual (non-zero) parameters nested within a larger distribution of total parameters of latent full-precision weights for a DNN block. Finally, the proposed PLUM framework achieves a 26% speedup on real hardware, doubles energy efficiency, and reduces density by 2.8x compared to binary methods while retaining top-1 accuracy when compared to prior-art methods for ResNets on ImageNet (by achieving 66.2% top-1 accuracy), presenting an alternative solution for deploying efficient models in resource-limited environments.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 29 of its 51 references that are in the map. First posted on arXiv 4 Dec 2023.

Builds on

29 of 51 references.

References · 29

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  4. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 7,005
  5. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,611
  6. 2016XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networkscited by 849
  7. 2021What Matters in Learning from Offline Human Demonstrations for Robot Manipulationcited by 725
  8. 2013Speech Recognition with Deep Recurrent Neural Networkscited by 669
  9. 2015BinaryConnect: Training Deep Neural Networks with binary weights during propagationscited by 592
  10. 2016DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradientscited by 569
  11. 2018LQ-Nets: Learned Quantization for Highly Accurate and Compact Deep Neural Networkscited by 274
  12. 2019Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networkscited by 238
17 more
  1. 2018Bi-Real Net: Enhancing the Performance of 1-bit CNNs With Improved Representational Capability and Advanced Training Algorithmcited by 235
  2. 2020SpAtten: Efficient Sparse Attention Architecture with Cascade Token and Head Pruningcited by 234
  3. 2016Trained Ternary Quantizationcited by 208
  4. 2016Ternary Weight Networkscited by 204
  5. 2019Forward and Backward Information Retention for Accurate Binary Neural Networkscited by 169
  6. 2017Towards Accurate Binary Convolutional Neural Networkcited by 149
  7. 2017Deep Learning with Low Precision by Half-wave Gaussian Quantizationcited by 117
  8. 2019XNOR-Net++: Improved Binary Neural Networkscited by 97
  9. 2018ProxQuant: Quantized Neural Networks via Proximal Operatorscited by 59
  10. 2021ReCU: Reviving the Dead Weights in Binary Neural Networkscited by 52
  11. 2021Distribution-sensitive Information Retention for Accurate Binary Neural Networkcited by 25
  12. 2018From Hashing to CNNs: Training BinaryWeight Networks via Hashingcited by 23
  13. 2019Improved training of binary networks for human pose estimation and image recognitioncited by 22
  14. 2020Least squares binary quantization of neural networkscited by 21
  15. 2018UCNN: Exploiting Computational Reuse in Deep Neural Networks via Weight Repetitioncited by 20
  16. 2020SparseTrain: Exploiting Dataflow Sparsity for Efficient Convolutional Neural Networks Trainingcited by 10
  17. 2023TeAAL: A Declarative Framework for Modeling Sparse Tensor Acceleratorscited by 4

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2023Performance Characterization of using Quantization for DNN Inference on Edge Devices: Extended Version
  2. 2024Effective Interplay between Sparsity and Quantization: From Theory to Practice
  3. 2021Enabling Binary Neural Network Training on the Edge
  4. 2021SparseDNN: Fast Sparse Deep Learning Inference on CPUs
  5. 2022REx: Data-Free Residual Quantization Error Expansion
  6. 2023DNA-TEQ: An Adaptive Exponential Quantization of Tensors for DNN Inference
  7. 2024Joint Pruning and Channel-wise Mixed-Precision Quantization for Efficient Deep Neural Networks
  8. 2021VS-Quant: Per-vector Scaled Quantization for Accurate Low-Precision Neural Network Inference