Data to 7 Oct 2026

A Multiscale Visualization of Attention in the Transformer Model

arXiv 1906.05714 ↗First posted 12 Jun 2019cs.HC cs.CL cs.LGcited by 292

Not on the map. Why?

Abstract (arXiv)

The Transformer is a sequence model that forgoes traditional recurrent architectures in favor of a fully attention-based approach. Besides improving performance, an advantage of using attention is that it can also help to interpret a model by showing how the model assigns weight to different input elements. However, the multi-layer, multi-head attention mechanism in the Transformer model can be difficult to decipher. To make the model more accessible, we introduce an open-source tool that visualizes attention at multiple scales, each of which provides a unique perspective on the attention mechanism. We demonstrate the tool on BERT and OpenAI GPT-2 and present three example use cases: detecting model bias, locating relevant attention heads, and linking neurons to model behavior.

Key facts. Cited by 292 papers in the map (data through 7 Oct 2026). First posted on arXiv 12 Jun 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 48; 2022: 56; 2023: 55; 2024: 50; 2025: 46; 2026: 37.

292 papers.

Sort:

By area:

2021GPT Understands, Toocited by 715
2023SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight Compressioncited by 359
2022Pre-Trained Language Models for Interactive Decision-Makingcited by 220
2021Synthesis of Compositional Animations from Textual Descriptionscited by 177
2024How do Large Language Models Handle Multilingualism?cited by 148
2021Understanding and Overcoming the Challenges of Efficient Transformer Quantizationcited by 139
2022Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networkscited by 126
2025Why do LLMs attend to the first token?cited by 122
2023SparQ Attention: Bandwidth-Efficient LLM Inferencecited by 119
2025HealthGPT: A Medical Large Vision-Language Model for Unifying Comprehension and Generation via Heterogeneous Knowledge Adaptationcited by 112
2023Circuit Component Reuse Across Tasks in Transformer Language Modelscited by 105
2021A Survey on Gender Bias in Natural Language Processingcited by 89

38 more

2021HiT: Hierarchical Transformer with Momentum Contrast for Video-Text Retrievalcited by 87
2024Trustworthiness in Retrieval-Augmented Generation Systems: A Surveycited by 83
2021BadPre: Task-agnostic Backdoor Attacks to Pre-trained NLP Foundation Modelscited by 76
2022Iteratively Prompt Pre-trained Language Models for Chain of Thoughtcited by 75
2024Zero Shot Health Trajectory Prediction Using Transformercited by 74
2023Copy Suppression: Comprehensively Understanding an Attention Headcited by 73
2022BiBERT: Accurate Fully Binarized BERTcited by 70
2022Interactive and Visual Prompt Engineering for Ad-hoc Task Adaptation with Large Language Modelscited by 70
2024Unveiling and Harnessing Hidden Attention Sinks: Enhancing Large Language Models without Training through Attention Calibrationcited by 70
2024On the Role of Attention Heads in Large Language Model Safetycited by 70
2022FormNet: Structural Encoding beyond Sequential Modeling in Form Document Information Extractioncited by 64
2021The Road to Know-Where: An Object-and-Room Informed Sequential BERT for Indoor Vision-Language Navigationcited by 60
2022Neural Media Bias Detection Using Distant Supervision With BABE -- Bias Annotations By Expertscited by 54
2023Why So Gullible? Enhancing the Robustness of Retrieval-Augmented Models against Counterfactual Noisecited by 52
2023AttentionViz: A Global View of Transformer Attentioncited by 50
2022Regression Transformer: Concurrent sequence regression and generation for molecular language modelingcited by 49
2022polyBERT: A chemical language model to enable fully machine-driven ultrafast polymer informaticscited by 48
2022Investigating Explainability of Generative AI for Code through Scenario-based Designcited by 47
2024Identifying Semantic Induction Heads to Understand In-Context Learningcited by 46
2022A Data-scalable Transformer for Medical Image Segmentation: Architecture, Model Efficiency, and Benchmarkcited by 45
2021CLiMP: A Benchmark for Chinese Language Model Evaluationcited by 44
2022Table Pre-training: A Survey on Model Architectures, Pre-training Objectives, and Downstream Taskscited by 42
2023Encoding Time-Series Explanations through Self-Supervised Model Behavior Consistencycited by 42
2024TokenSHAP: Interpreting Large Language Models with Monte Carlo Shapley Value Estimationcited by 41
2022Spatiality-guided Transformer for 3D Dense Captioning on Point Cloudscited by 40
2022Analyzing Encoded Concepts in Transformer Language Modelscited by 37
2024A Cross-Language Investigation into Jailbreak Attacks in Large Language Modelscited by 37
2023Understanding Addition in Transformerscited by 34
2022LISA: Learning Interpretable Skill Abstractions from Languagecited by 30
2023How Does Attention Work in Vision Transformers? A Visual Analytics Attemptcited by 29
2021Dodrio: Exploring Transformer Models with Interactive Visualizationcited by 28
2022MOFormer: Self-Supervised Transformer model for Metal-Organic Framework Property Predictioncited by 28
2022Interacting with next-phrase suggestions: How suggestion systems aid and influence the cognitive processes of writingcited by 27
2021Accounting for Agreement Phenomena in Sentence Comprehension with Transformer Language Models: Effects of Similarity-based Interference on Surprisal and Attentioncited by 26
2024On Behalf of the Stakeholders: Trends in NLP Model Interpretability in the Era of LLMscited by 24
2023Skip-Attention: Improving Vision Transformers by Paying Less Attentioncited by 23
2024XAI meets LLMs: A Survey of the Relation between Explainable AI and Large Language Modelscited by 23
2021Relative Importance in Sentence Processingcited by 22

50 most cited of 292 All 292 (JSON)