Efficient VQ-QAT and Mixed Vector/Linear quantized Neural Networks

• URL: http://arxiv.org/abs/2604.23172v1
• Date: Sat, 25 Apr 2026 06:55:04 GMT
• Title: Efficient VQ-QAT and Mixed Vector/Linear quantized Neural Networks
• Abstract summary: In this work, we developed and tested 3 techniques for vector quantization (VQ) based model weight compression.<n>To mitigate codebook collapse and enable end-to-end training, we adopted cosine similarity-based assignment.<n>We also investigated the use of differentiable neural architecture search (NAS) to adaptively select layer-wise quantization configurations.
• Score: 1.8916657412668536
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In this work, we developed and tested 3 techniques for vector quantization (VQ) based model weight compression. To mitigate codebook collapse and enable end-to-end training, we adopted cosine similarity-based assignment. Building on ideas from attention-based formulations in Differentiable K-Means (DKM), we further improved this approach by using cosine similarity for assignment combined with top-1 sampling and a straight-through estimator, thereby eliminating the need for weighted-average reconstruction. Finally, we investigated the use of differentiable neural architecture search (NAS) to adaptively select layer-wise quantization configurations, further optimizing the compression process. Although our method does not consistently outperform existing approaches across all quantization levels, it provides useful insights into the design trade-offs and behaviors of VQ-based model compression methods.

Related papers

• Rethinking Autoregressive Models for Lossless Image Compression via Hierarchical Parallelism and Progressive Adaptation [75.58269386927076]
  Autoregressive (AR) models are often dismissed as impractical due to prohibitive computational cost.<n>This work re-thinks this paradigm, introducing a framework built on hierarchical parallelism and progressive adaptation.<n> Experiments on diverse datasets (natural, satellite, medical) validate that our method achieves new state-of-the-art compression.
  arXiv  Detail & Related papers  (2025-11-14T06:27:58Z)
• Preventing Local Pitfalls in Vector Quantization via Optimal Transport [77.15924044466976]
  We introduce OptVQ, a novel vector quantization method that employs the Sinkhorn algorithm to optimize the optimal transport problem.<n>Our experiments on image reconstruction tasks demonstrate that OptVQ achieves 100% codebook utilization and surpasses current state-of-the-art VQNs in reconstruction quality.
  arXiv  Detail & Related papers  (2024-12-19T18:58:14Z)
• Learning Optimal Lattice Vector Quantizers for End-to-end Neural Image Compression [16.892815659154053]
  Lattice vector quantization (LVQ) presents a compelling alternative, which can exploit inter-feature dependencies more effectively. Traditional LVQ structures are designed/optimized for uniform source distributions. We propose a novel learning method to overcome this weakness by designing the rate-distortion optimal lattice vector quantization codebooks.
  arXiv  Detail & Related papers  (2024-11-25T06:05:08Z)
• Pointer Networks with Q-Learning for Combinatorial Optimization [55.2480439325792]
  We introduce the Pointer Q-Network (PQN), a hybrid neural architecture that integrates model-free Q-value policy approximation with Pointer Networks (Ptr-Nets) Our empirical results demonstrate the efficacy of this approach, also testing the model in unstable environments.
  arXiv  Detail & Related papers  (2023-11-05T12:03:58Z)
• QuantEase: Optimization-based Quantization for Language Models [17.333778751252392]
  This work introduces Quantization (PTQ) of various quantization layers from recent advances of Large Language Models (LLMs) Our CD-based approach features straightforward updates, relying solely on vector operations. We also explore an outlier approach, allowing for retaining significant weights (outoutliers) with complete precision.
  arXiv  Detail & Related papers  (2023-09-05T01:39:09Z)
• Quantization Aware Factorization for Deep Neural Network Compression [20.04951101799232]
  decomposition of convolutional and fully-connected layers is an effective way to reduce parameters and FLOP in neural networks. A conventional post-training quantization approach applied to networks with weights yields a drop in accuracy. This motivated us to develop an algorithm that finds decomposed approximation directly with quantized factors.
  arXiv  Detail & Related papers  (2023-08-08T21:38:02Z)
• Weight Re-Mapping for Variational Quantum Algorithms [54.854986762287126]
  We introduce the concept of weight re-mapping for variational quantum circuits (VQCs) We employ seven distinct weight re-mapping functions to assess their impact on eight classification datasets. Our results indicate that weight re-mapping can enhance the convergence speed of the VQC.
  arXiv  Detail & Related papers  (2023-06-09T09:42:21Z)
• LVQAC: Lattice Vector Quantization Coupled with Spatially Adaptive Companding for Efficient Learned Image Compression [24.812267280543693]
  We present a novel Lattice Vector Quantization scheme coupled with a spatially Adaptive Companding (LVQAC) mapping. For any end-to-end CNN image compression models, replacing uniform quantizer by LVQAC achieves better rate-distortion performance without significantly increasing the model complexity.
  arXiv  Detail & Related papers  (2023-03-25T23:34:15Z)
• Reducing Computational Complexity of Neural Networks in Optical Channel Equalization: From Concepts to Implementation [1.6987798749419218]
  We show that it is possible to design an NN-based equalizer that is simpler to implement and has better performance than the conventional digital back-propagation (DBP) equalizer with only one step per span. An equalizer based on NN can also achieve superior performance while still maintaining the same degree of complexity as the full electronic chromatic dispersion compensation block.
  arXiv  Detail & Related papers  (2022-08-26T21:00:05Z)
• Learning Representations for CSI Adaptive Quantization and Feedback [51.14360605938647]
  We propose an efficient method for adaptive quantization and feedback in frequency division duplexing systems. Existing works mainly focus on the implementation of autoencoder (AE) neural networks for CSI compression. We recommend two different methods: one based on a post training quantization and the second one in which the codebook is found during the training of the AE.
  arXiv  Detail & Related papers  (2022-07-13T08:52:13Z)
• BSQ: Exploring Bit-Level Sparsity for Mixed-Precision Neural Network Quantization [32.770842274996774]
  Mixed-precision quantization can potentially achieve the optimal tradeoff between performance and compression rate of deep neural networks. Previous methods either examine only a small manually-designed search space or utilize a cumbersome neural architecture search to explore the vast search space. This work proposes bit-level sparsity quantization (BSQ) to tackle the mixed-precision quantization from a new angle of inducing bit-level sparsity.
  arXiv  Detail & Related papers  (2021-02-20T22:37:41Z)
• Once Quantization-Aware Training: High Performance Extremely Low-bit Architecture Search [112.05977301976613]
  We propose to combine Network Architecture Search methods with quantization to enjoy the merits of the two sides. We first propose the joint training of architecture and quantization with a shared step size to acquire a large number of quantized models. Then a bit-inheritance scheme is introduced to transfer the quantized models to the lower bit, which further reduces the time cost and improves the quantization accuracy.
  arXiv  Detail & Related papers  (2020-10-09T03:52:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.