TransforLearn: Interactive Visual Tutorial for the Transformer Model

Lin Gao, Zekai Shao, Ziqin LUO, Haibo Hu, Cagatay Turkay, Siming Chen

DOI: 10.1109/TVCG.2023.3327353

Room: 109

2023-10-25T22:00:00ZGMT-0600Change your timezone on the schedule page
2023-10-25T22:00:00Z
Exemplar figure, described by caption below
With TransforLearn, learners can gain an understanding of the Transformer structure and the process of machine translation. Input view (A) provides an interface for the text to be translated. Translation view (B) displays the translation results and current translation progress, helping users in task-driven exploration. Architecture view (C) provides an overview of model structure and data flow, with sub-views (C1-C4) that support computational processes. Once enabled, the Detailed view (C3) displays the Attention mechanism view (D), Layer normalization view (E), and Feed-forward network view (F). These views not only show the operational details but also support multiple interactions.
Fast forward
Full Video
Keywords

Deep learning, Transformer, Visual tutorial, Explorable explanations

Abstract

The widespread adoption of Transformers in deep learning, serving as the core framework for numerous large-scale language models, has sparked significant interest in understanding their underlying mechanisms. However, beginners face difficulties in comprehending and learning Transformers due to its complex structure and abstract data representation. We present TransforLearn, the first interactive visual tutorial designed for deep learning beginners and non-experts to comprehensively learn about Transformers. TransforLearn supports interactions for architecture-driven exploration and task-driven exploration, providing insight into different levels of model details and their working processes. It accommodates interactive views of each layer's operation and mathematical formula, helping users to understand the data flow of long text sequences. By altering the current decoder-based recursive prediction results and combining the downstream task abstractions, users can deeply explore model processes. Our user study revealed that the interactions of TransforLearn are positively received. We observe that TransforLearn facilitates users' accomplishment of study tasks and a grasp of key concepts in Transformer effectively.