# Interpreting Deep Visual Representations via Network Dissection

arXiv: https://arxiv.org/abs/1711.05611 · HTML: https://mlsanity.com/paper/1711.05611/

## Key facts

- Cited by 103 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Nov 2017.

## Built on by (most cited first, 30 of 103)

- [Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges](https://mlsanity.com/paper/2103.11251/) (2021)
- [Generic Attention-model Explainability for Interpreting Bi-Modal and Encoder-Decoder Transformers](https://mlsanity.com/paper/2103.15679/) (2021)
- [Does Localization Inform Editing? Surprising Differences in Causality-Based Localization vs. Knowledge Editing in Language Models](https://mlsanity.com/paper/2301.04213/) (2023)
- [Patchscopes: A Unifying Framework for Inspecting Hidden Representations of Language Models](https://mlsanity.com/paper/2401.06102/) (2024)
- [Salient ImageNet: How to discover spurious features in Deep Learning?](https://mlsanity.com/paper/2110.04301/) (2021)
- [Weakly-Supervised Multi-Granularity Map Learning for Vision-and-Language Navigation](https://mlsanity.com/paper/2210.07506/) (2022)
- [Neuro-Symbolic Artificial Intelligence: Current Trends](https://mlsanity.com/paper/2105.05330/) (2021)
- [Explainable and Interpretable Multimodal Large Language Models: A Comprehensive Survey](https://mlsanity.com/paper/2412.02104/) (2024)
- [ACAV100M: Automatic Curation of Large-Scale Datasets for Audio-Visual Video Representation Learning](https://mlsanity.com/paper/2101.10803/) (2021)
- [Multimodal Neurons in Pretrained Text-Only Transformers](https://mlsanity.com/paper/2308.01544/) (2023)
- [LINe: Out-of-Distribution Detection by Leveraging Important Neurons](https://mlsanity.com/paper/2303.13995/) (2023)
- [Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces](https://mlsanity.com/paper/2212.14855/) (2022)
- [Interpretable and Testable Vision Features via Sparse Autoencoders](https://mlsanity.com/paper/2502.06755/) (2025)
- [FGPrompt: Fine-grained Goal Prompting for Image-goal Navigation](https://mlsanity.com/paper/2310.07473/) (2023)
- [Capturing the objects of vision with neural networks](https://mlsanity.com/paper/2109.03351/) (2021)
- [Improving Molecular Graph Neural Network Explainability with Orthonormalization and Induced Sparsity](https://mlsanity.com/paper/2105.04854/) (2021)
- [Attribute Prototype Network for Any-Shot Learning](https://mlsanity.com/paper/2204.01208/) (2022)
- [Visual Explanations via Iterated Integrated Attributions](https://mlsanity.com/paper/2310.18585/) (2023)
- [Token Transformation Matters: Towards Faithful Post-hoc Explanation for Vision Transformer](https://mlsanity.com/paper/2403.14552/) (2024)
- [On the Faithfulness of Vision Transformer Explanations](https://mlsanity.com/paper/2404.01415/) (2024)
- [HINT: Hierarchical Neuron Concept Explainer](https://mlsanity.com/paper/2203.14196/) (2022)
- [Revealing Hidden Context Bias in Segmentation and Object Detection through Concept-specific Explanations](https://mlsanity.com/paper/2211.11426/) (2022)
- [Core Risk Minimization using Salient ImageNet](https://mlsanity.com/paper/2203.15566/) (2022)
- [A Simple Interpretable Transformer for Fine-Grained Image Classification and Analysis](https://mlsanity.com/paper/2311.04157/) (2023)
- [Mechanistic Finetuning of Vision-Language-Action Models via Few-Shot Demonstrations](https://mlsanity.com/paper/2511.22697/) (2025)
- [Networks are Slacking Off: Understanding Generalization Problem in Image Deraining](https://mlsanity.com/paper/2305.15134/) (2023)
- [Quality-Aware Network for Human Parsing](https://mlsanity.com/paper/2103.05997/) (2021)
- [Explaining Deep Convolutional Neural Networks via Latent Visual-Semantic Filter Attention](https://mlsanity.com/paper/2204.04601/) (2022)
- [MIXRTs: Toward Interpretable Multi-Agent Reinforcement Learning via Mixing Recurrent Soft Decision Trees](https://mlsanity.com/paper/2209.07225/) (2022)
- [New Definitions and Evaluations for Saliency Methods: Staying Intrinsic, Complete and Sound](https://mlsanity.com/paper/2211.02912/) (2022)

## Abstract (arXiv)

The success of recent deep convolutional neural networks (CNNs) depends on learning hidden representations that can summarize the important factors of variation behind the data. However, CNNs often criticized as being black boxes that lack interpretability, since they have millions of unexplained model parameters. In this work, we describe Network Dissection, a method that interprets networks by providing labels for the units of their deep visual representations. The proposed method quantifies the interpretability of CNN representations by evaluating the alignment between individual hidden units and a set of visual semantic concepts. By identifying the best alignments, units are given human interpretable labels across a range of objects, parts, scenes, textures, materials, and colors. The method reveals that deep representations are more transparent and interpretable than expected: we find that representations are significantly more interpretable than they would be under a random equivalently powerful basis. We apply the method to interpret and compare the latent representations of various network architectures trained to solve different supervised and self-supervised training tasks. We then examine factors affecting the network interpretability such as the number of the training iterations, regularizations, different initializations, and the network depth and width. Finally we show that the interpreted units can be used to provide explicit explanations of a prediction given by a CNN for an image. Our results highlight that interpretability is an important property of deep neural networks that provides new insights into their hierarchical structure.