# SYQ: Learning Symmetric Quantization For Efficient Deep Neural Networks

arXiv: https://arxiv.org/abs/1807.00301 · HTML: https://mlsanity.com/paper/1807.00301/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Jul 2018.

## Built on by (most cited first, 24 of 24)

- [A Survey of Quantization Methods for Efficient Neural Network Inference](https://mlsanity.com/paper/2103.13630/) (2021)
- [Pruning and Quantization for Deep Neural Network Acceleration: A Survey](https://mlsanity.com/paper/2101.09671/) (2021)
- [BiLLM: Pushing the Limit of Post-Training Quantization for LLMs](https://mlsanity.com/paper/2402.04291/) (2024)
- [A Comprehensive Survey of Dataset Distillation](https://mlsanity.com/paper/2301.05603/) (2023)
- [It's All In the Teacher: Zero-Shot Quantization Brought Closer to the Teacher](https://mlsanity.com/paper/2203.17008/) (2022)
- [LLMEasyQuant: Scalable Quantization for Parallel and Distributed LLM Inference](https://mlsanity.com/paper/2406.19657/) (2024)
- [LilNetX: Lightweight Networks with EXtreme Model Compression and Structured Sparsification](https://mlsanity.com/paper/2204.02965/) (2022)
- [Mixed-Precision Neural Networks: A Survey](https://mlsanity.com/paper/2208.06064/) (2022)
- [FAT: Learning Low-Bitwidth Parametric Representation via Frequency-Aware Transformation](https://mlsanity.com/paper/2102.07444/) (2021)
- [Bi-VLM: Pushing Ultra-Low Precision Post-Training Quantization Boundaries in Vision-Language Models](https://mlsanity.com/paper/2509.18763/) (2025)
- [Elastic Significant Bit Quantization and Acceleration for Deep Neural Networks](https://mlsanity.com/paper/2109.03513/) (2021)
- [ARC: A Vision-based Automatic Retail Checkout System](https://mlsanity.com/paper/2104.02832/) (2021)
- [LDP: Learnable Dynamic Precision for Efficient Deep Neural Network Training and Inference](https://mlsanity.com/paper/2203.07713/) (2022)
- [FxP-QNet: A Post-Training Quantizer for the Design of Mixed Low-Precision DNNs with Dynamic Fixed-Point Representation](https://mlsanity.com/paper/2203.12091/) (2022)
- [AskewSGD : An Annealed interval-constrained Optimisation method to train Quantized Neural Networks](https://mlsanity.com/paper/2211.03741/) (2022)
- [Ternary Quantization: A Survey](https://mlsanity.com/paper/2303.01505/) (2023)
- [From Algorithm to Hardware: A Survey on Efficient and Safe Deployment of Deep Neural Networks](https://mlsanity.com/paper/2405.06038/) (2024)
- [Compensate Quantization Errors: Make Weights Hierarchical to Compensate Each Other](https://mlsanity.com/paper/2406.16299/) (2024)
- [RedBit: An End-to-End Flexible Framework for Evaluating the Accuracy of Quantized CNNs](https://mlsanity.com/paper/2301.06193/) (2023)
- [MobileNMT: Enabling Translation in 15MB and 30ms](https://mlsanity.com/paper/2306.04235/) (2023)
- [Quantized Neural Networks via {-1, +1} Encoding Decomposition and Acceleration](https://mlsanity.com/paper/2106.09886/) (2021)
- [Learning Discrete Weights and Activations Using the Local Reparameterization Trick](https://mlsanity.com/paper/2307.01683/) (2023)
- [NASH: Neural Architecture Search for Hardware-Optimized Machine Learning Models](https://mlsanity.com/paper/2403.01845/) (2024)
- [Compensate Quantization Errors+: Quantized Models Are Inquisitive Learners](https://mlsanity.com/paper/2407.15508/) (2024)

## Abstract (arXiv)

Inference for state-of-the-art deep neural networks is computationally expensive, making them difficult to deploy on constrained hardware environments. An efficient way to reduce this complexity is to quantize the weight parameters and/or activations during training by approximating their distributions with a limited entry codebook. For very low-precisions, such as binary or ternary networks with 1-8-bit activations, the information loss from quantization leads to significant accuracy degradation due to large gradient mismatches between the forward and backward functions. In this paper, we introduce a quantization method to reduce this loss by learning a symmetric codebook for particular weight subgroups. These subgroups are determined based on their locality in the weight matrix, such that the hardware simplicity of the low-precision representations is preserved. Empirically, we show that symmetric quantization can substantially improve accuracy for networks with extremely low-precision weights and activations. We also demonstrate that this representation imposes minimal or no hardware implications to more coarse-grained approaches. Source code is available at https://www.github.com/julianfaraone/SYQ.