# Visually Analyzing Contextualized Embeddings

arXiv: https://arxiv.org/abs/2009.02554 · HTML: https://mlsanity.com/paper/2009.02554/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Sep 2020.

## Built on by (most cited first, 7 of 7)

- [Visual Analytics For Machine Learning: A Data Perspective Survey](https://mlsanity.com/paper/2307.07712/) (2023)
- [Visual Comparison of Language Model Adaptation](https://mlsanity.com/paper/2208.08176/) (2022)
- [SensePOLAR: Word sense aware interpretability for pre-trained contextual word embeddings](https://mlsanity.com/paper/2301.04704/) (2023)
- [Explainable Mapper: Charting LLM Embedding Spaces Using Perturbation-Based Explanation and Verification Agents](https://mlsanity.com/paper/2507.18607/) (2025)
- [EmbeddingTree: Hierarchical Exploration of Entity Features in Embedding](https://mlsanity.com/paper/2308.01329/) (2023)
- [LayerFlow: Layer-wise Exploration of LLM Embeddings using Uncertainty-aware Interlinked Projections](https://mlsanity.com/paper/2504.10504/) (2025)
- [Visualising Information Flow in Word Embeddings with Diffusion Tensor Imaging](https://mlsanity.com/paper/2601.05713/) (2026)

## Abstract (arXiv)

In this paper we introduce a method for visually analyzing contextualized embeddings produced by deep neural network-based language models. Our approach is inspired by linguistic probes for natural language processing, where tasks are designed to probe language models for linguistic structure, such as parts-of-speech and named entities. These approaches are largely confirmatory, however, only enabling a user to test for information known a priori. In this work, we eschew supervised probing tasks, and advocate for unsupervised probes, coupled with visual exploration techniques, to assess what is learned by language models. Specifically, we cluster contextualized embeddings produced from a large text corpus, and introduce a visualization design based on this clustering and textual structure - cluster co-occurrences, cluster spans, and cluster-word membership - to help elicit the functionality of, and relationship between, individual clusters. User feedback highlights the benefits of our design in discovering different types of linguistic structures.