[Feature] Proposal: Chinese (zh-CN) localization for Transformer Explainer

Author: Jian1202Created Dec 14, 2025Updated Jul 24, 2026

Hi!

First of all, thank you for creating Transformer Explainer. It’s one of the most intuitive tools I’ve seen for building concrete intuition about Transformer internals, especially attention and residual connections.

I’d like to ask whether you would be open to adding an optional Chinese (zh-CN) localization for the UI and explanatory text.

About my perspective:

I’m currently an undergraduate CS student, and I often see tools like Transformer Explainer used in real learning settings among Chinese-speaking students. From that vantage point, a recurring friction shows up: the visual explanations themselves are usually easy to follow, but English UI labels and explanatory text can interrupt the “aha” moment for learners who are otherwise tracking the visualization just fine.

This observation is what motivated me to work on a Chinese (zh-CN) localization. The goal is not to change or reinterpret the content, but simply to remove the language barrier so learners can engage more directly with the existing visual explanations.

What I’ve done:

I’ve completed a working Chinese localization that includes:

  • UI labels
  • Tooltips
  • Explanatory text shown alongside the visualizations

All model logic, interaction behavior, and visualization mechanics are kept exactly the same. The changes are strictly limited to text and UI presentation.

To be explicit: this was not a class assignment.

I worked on this because the language barrier is the primary factor slowing down adoption and effective use in my local learning context, and localization is a straightforward way to address that without touching the core implementation.

If this aligns with the project’s goals, I’d be happy to submit a PR and adapt the structure (e.g. i18n layout, file organization, naming) to match whatever approach you’d prefer for maintaining translations going forward.

For reference, here is the current working version:

Thanks again for building such a great educational tool. I’d really appreciate any feedback on whether this direction makes sense for the project, and how you’d like me to proceed.

Source: poloclub/transformer-explainer