# Class-constrained t-SNE: Combining Data Features and Class Probabilities

arXiv: https://arxiv.org/abs/2308.13837 · HTML: https://mlsanity.com/paper/2308.13837/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 61 references that are in the map.
- First posted on arXiv 26 Aug 2023.

## Builds on (most cited first, 10 of 10)

- [Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning](https://mlsanity.com/paper/1506.02142/) (2015)
- [Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms](https://mlsanity.com/paper/1708.07747/) (2017)
- [Distributed Representations of Sentences and Documents](https://mlsanity.com/paper/1405.4053/) (2014)
- [Manifold: A Model-Agnostic Framework for Interpretation and Diagnosis of Machine Learning Models](https://mlsanity.com/paper/1808.00196/) (2018)
- [Attraction-Repulsion Spectrum in Neighbor Embeddings](https://mlsanity.com/paper/2007.08902/) (2020)
- [Embedding Comparator: Visualizing Differences in Global Structure and Local Neighborhoods via Small Multiples](https://mlsanity.com/paper/1912.04853/) (2019)
- [Boxer: Interactive Comparison of Classifier Results](https://mlsanity.com/paper/2004.07964/) (2020)
- [Explaining Vulnerabilities to Adversarial Machine Learning through Visual Analytics](https://mlsanity.com/paper/1907.07296/) (2019)
- [DFSeer: A Visual Analytics Approach to Facilitate Model Selection for Demand Forecasting](https://mlsanity.com/paper/2005.03244/) (2020)
- [Visual Integration of Data and Model Space in Ensemble Learning](https://mlsanity.com/paper/1710.07322/) (2017)

## Built on by (most cited first, 8 of 8)

- [Stop Misusing t-SNE and UMAP for Visual Analytics](https://mlsanity.com/paper/2506.08725/) (2025)
- [EvolvED: Evolutionary Embeddings to Understand the Generation Process of Diffusion Models](https://mlsanity.com/paper/2406.17462/) (2024)
- [CUPID: Contextual Understanding of Prompt-conditioned Image Distributions](https://mlsanity.com/paper/2406.07699/) (2024)
- [Injecting Explainability and Lightweight Design into Weakly Supervised Video Anomaly Detection Systems](https://mlsanity.com/paper/2412.20201/) (2024)
- [Why Can't I See My Clusters? A Precision-Recall Approach to Dimensionality Reduction Validation](https://mlsanity.com/paper/2509.04222/) (2025)
- [A Multi-Attribute Latent Space for Visual Analysis of Watches](https://mlsanity.com/paper/2606.27897/) (2026)
- [Scalable Semantic Steering of Embedding Projections](https://mlsanity.com/paper/2607.03978/) (2026)
- [Human-Guided Causal Knowledge Injection for Virtual Cells](https://mlsanity.com/paper/2608.08430/) (2026)

## Abstract (arXiv)

Data features and class probabilities are two main perspectives when, e.g., evaluating model results and identifying problematic items. Class probabilities represent the likelihood that each instance belongs to a particular class, which can be produced by probabilistic classifiers or even human labeling with uncertainty. Since both perspectives are multi-dimensional data, dimensionality reduction (DR) techniques are commonly used to extract informative characteristics from them. However, existing methods either focus solely on the data feature perspective or rely on class probability estimates to guide the DR process. In contrast to previous work where separate views are linked to conduct the analysis, we propose a novel approach, class-constrained t-SNE, that combines data features and class probabilities in the same DR result. Specifically, we combine them by balancing two corresponding components in a cost function to optimize the positions of data points and iconic representation of classes -- class landmarks. Furthermore, an interactive user-adjustable parameter balances these two components so that users can focus on the weighted perspectives of interest and also empowers a smooth visual transition between varying perspectives to preserve the mental map. We illustrate its application potential in model evaluation and visual-interactive labeling. A comparative analysis is performed to evaluate the DR results.