# Probing neural networks with t-SNE, class-specific projections and a guided tour

arXiv: https://arxiv.org/abs/2107.12547 · HTML: https://mlsanity.com/paper/2107.12547/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 17 references that are in the map.
- First posted on arXiv 27 Jul 2021.

## Builds on (most cited first, 4 of 4)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Understanding intermediate layers using linear classifier probes](https://mlsanity.com/paper/1610.01644/) (2016)
- [Prevalence of Neural Collapse during the terminal phase of deep learning training](https://mlsanity.com/paper/2008.08186/) (2020)
- [Do Wide and Deep Networks Learn the Same Things? Uncovering How Neural Network Representations Vary with Width and Depth](https://mlsanity.com/paper/2010.15327/) (2020)

## Built on by (most cited first, 6 of 6)

- [Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networks](https://mlsanity.com/paper/2207.13243/) (2022)
- [Linguistic Collapse: Neural Collapse in (Large) Language Models](https://mlsanity.com/paper/2405.17767/) (2024)
- [Transformer Block Coupling and its Correlation with Generalization in LLMs](https://mlsanity.com/paper/2407.07810/) (2024)
- [A Plot is Worth a Thousand Words: Model Information Stealing Attacks via Scientific Plots](https://mlsanity.com/paper/2302.11982/) (2023)
- [Sentence Ambiguity, Grammaticality and Complexity Probes](https://mlsanity.com/paper/2210.06928/) (2022)
- [Gradient Smoothing: Coupling Layer-wise Updates for Improved Optimization](https://mlsanity.com/paper/2606.30813/) (2026)

## Abstract (arXiv)

We use graphical methods to probe neural nets that classify images. Plots of t-SNE outputs at successive layers in a network reveal increasingly organized arrangement of the data points. They can also reveal how a network can diminish or even forget about within-class structure as the data proceeds through layers. We use class-specific analogues of principal components to visualize how succeeding layers separate the classes. These allow us to sort images from a given class from most typical to least typical (in the data) and they also serve as very useful projection coordinates for data visualization. We find them especially useful when defining versions guided tours for animated data visualization.