# Bridging the Accuracy Gap for 2-bit Quantized Neural Networks (QNN)

arXiv: https://arxiv.org/abs/1807.06964 · HTML: https://mlsanity.com/paper/1807.06964/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Jul 2018.

## Built on by (most cited first, 19 of 19)

- [CBQ: Cross-Block Quantization for Large Language Models](https://mlsanity.com/paper/2312.07950/) (2023)
- [Accurate Neural Training with 4-bit Matrix Multiplications at Standard Formats](https://mlsanity.com/paper/2112.10769/) (2021)
- [Quantization and Deployment of Deep Neural Networks on Microcontrollers](https://mlsanity.com/paper/2105.13331/) (2021)
- [Differentiable Dynamic Quantization with Mixed Precision and Adaptive Resolution](https://mlsanity.com/paper/2106.02295/) (2021)
- [2DQuant: Low-bit Post-Training Quantization for Image Super-Resolution](https://mlsanity.com/paper/2406.06649/) (2024)
- [Minimum Variance Unbiased N:M Sparsity for the Neural Gradients](https://mlsanity.com/paper/2203.10991/) (2022)
- [4-bit Quantization of LSTM-based Speech Recognition Models](https://mlsanity.com/paper/2108.12074/) (2021)
- [DeepGEMM: Accelerated Ultra Low-Precision Inference on CPU Architectures using Lookup Tables](https://mlsanity.com/paper/2304.09049/) (2023)
- [Emergent Quantized Communication](https://mlsanity.com/paper/2211.02412/) (2022)
- [Finding Interpretable Class-Specific Patterns through Efficient Neural Search](https://mlsanity.com/paper/2312.04311/) (2023)
- [Outlier-Aware Post-Training Quantization for Image Super-Resolution](https://mlsanity.com/paper/2511.00682/) (2025)
- [BOLD: Boolean Logic Deep Learning](https://mlsanity.com/paper/2405.16339/) (2024)
- [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)
- [Latency-Constrained DNN Architecture Learning for Edge Systems using Zerorized Batch Normalization](https://mlsanity.com/paper/2607.06922/) (2026)
- [DNN Quantization with Attention](https://mlsanity.com/paper/2103.13322/) (2021)
- [Realizing Unaligned Block-wise Pruning for DNN Acceleration on Mobile Devices](https://mlsanity.com/paper/2407.19644/) (2024)
- [Rethinking Output Alignment For 1-bit Post-Training Quantization of Large Language Models](https://mlsanity.com/paper/2512.21651/) (2025)
- [StatQAT: Statistical Quantizer Optimization for Deep Networks](https://mlsanity.com/paper/2605.17745/) (2026)

## Abstract (arXiv)

Deep learning algorithms achieve high classification accuracy at the expense of significant computation cost. In order to reduce this cost, several quantization schemes have gained attention recently with some focusing on weight quantization, and others focusing on quantizing activations. This paper proposes novel techniques that target weight and activation quantizations separately resulting in an overall quantized neural network (QNN). The activation quantization technique, PArameterized Clipping acTivation (PACT), uses an activation clipping parameter α that is optimized during training to find the right quantization scale. The weight quantization scheme, statistics-aware weight binning (SAWB), finds the optimal scaling factor that minimizes the quantization error based on the statistical characteristics of the distribution of weights without the need for an exhaustive search. The combination of PACT and SAWB results in a 2-bit QNN that achieves state-of-the-art classification accuracy (comparable to full precision networks) across a range of popular models and datasets.