# Entanglement and Tensor Networks for Supervised Image Classification

arXiv: https://arxiv.org/abs/2007.06082 · HTML: https://mlsanity.com/paper/2007.06082/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Jul 2020.

## Built on by (most cited first, 18 of 18)

- [Grokking phase transitions in learning local rules with gradient descent](https://mlsanity.com/paper/2210.15435/) (2022)
- [Generative modeling with projected entangled-pair states](https://mlsanity.com/paper/2202.08177/) (2022)
- [Machine learning with tree tensor networks, CP rank constraints, and tensor dropout](https://mlsanity.com/paper/2305.19440/) (2023)
- [DISP6D: Disentangled Implicit Shape and Pose Learning for Scalable 6D Pose Estimation](https://mlsanity.com/paper/2107.12549/) (2021)
- [Quantum-inspired event reconstruction with Tensor Networks: Matrix Product States](https://mlsanity.com/paper/2106.08334/) (2021)
- [Generative Learning of Continuous Data by Tensor Networks](https://mlsanity.com/paper/2310.20498/) (2023)
- [Classical versus Quantum: comparing Tensor Network-based Quantum Circuits on LHC data](https://mlsanity.com/paper/2202.10471/) (2022)
- [TensorKrowch: Smooth integration of tensor networks in machine learning](https://mlsanity.com/paper/2306.08595/) (2023)
- [tn4ml: Tensor Network Training and Customization for Machine Learning](https://mlsanity.com/paper/2502.13090/) (2025)
- [ATP: Adaptive Threshold Pruning for Efficient Data Encoding in Quantum Neural Networks](https://mlsanity.com/paper/2503.21815/) (2025)
- [Tensor network to learn the wavefunction of data](https://mlsanity.com/paper/2111.08014/) (2021)
- [What Makes Data Suitable for a Locally Connected Neural Network? A Necessary and Sufficient Condition Based on Quantum Entanglement](https://mlsanity.com/paper/2303.11249/) (2023)
- [Generalization and Overfitting in Matrix Product State Machine Learning Architectures](https://mlsanity.com/paper/2208.04372/) (2022)
- [Differentiable Programming of Isometric Tensor Networks](https://mlsanity.com/paper/2110.03898/) (2021)
- [Deep tensor networks with matrix product operators](https://mlsanity.com/paper/2209.09098/) (2022)
- [Distributive Pre-Training of Generative Modeling Using Matrix-Product States](https://mlsanity.com/paper/2306.14787/) (2023)
- [Using matrix-product states for time-series machine learning](https://mlsanity.com/paper/2412.15826/) (2024)
- [Entangling Machine Learning with Quantum Tensor Networks](https://mlsanity.com/paper/2403.12969/) (2024)

## Abstract (arXiv)

Tensor networks, originally designed to address computational problems in quantum many-body physics, have recently been applied to machine learning tasks. However, compared to quantum physics, where the reasons for the success of tensor network approaches over the last 30 years is well understood, very little is yet known about why these techniques work for machine learning. The goal of this paper is to investigate entanglement properties of tensor network models in a current machine learning application, in order to uncover general principles that may guide future developments. We revisit the use of tensor networks for supervised image classification using the MNIST data set of handwritten digits, as pioneered by Stoudenmire and Schwab [Adv. in Neur. Inform. Proc. Sys. 29, 4799 (2016)]. Firstly we hypothesize about which state the tensor network might be learning during training. For that purpose, we propose a plausible candidate state |Sigmaℓrangle (built as a superposition of product states corresponding to images in the training set) and investigate its entanglement properties. We conclude that |Sigmaℓrangle is so robustly entangled that it cannot be approximated by the tensor network used in that work, which must therefore be representing a very different state. Secondly, we use tensor networks with a block product structure, in which entanglement is restricted within small blocks of n × n pixels/qubits. We find that these states are extremely expressive (e.g. training accuracy of 99.97 % already for n=2), suggesting that long-range entanglement may not be essential for image classification. However, in our current implementation, optimization leads to over-fitting, resulting in test accuracies that are not competitive with other current approaches.