2026 · Conference paper

Dashboard2Code: Evaluating Multimodal Models on Reconstructing Interactive Dashboards

Evaluates agents that actively explore a running dashboard and reconstruct its appearance and interaction logic. DashboardMimic provides 180 Plotly+Dash examples, with code-semantic analysis and dynamic interaction tests for the generated implementations.

Scope and limitations

Limited to Plotly+Dash dashboards, including model-augmented examples. Some settings expose DOM information and evaluation uses model judges. This is not a general backend-recovery benchmark. The linked dataset requires accepting access conditions and sharing contact information.

Why it fits this collection

Directly studies reconstruction through active observation, including callbacks and state dependencies rather than screenshot appearance alone.

Publication and review details

Venue
ACL 2026
Status
Conference paper
Publication / posting date
July 2026
Date source
ACL proceedings publication month.
First arXiv submission
July 6, 2026
arXiv
2607.04727
Reviewed arXiv version
v1
DOI
10.18653/v1/2026.acl-long.1750

The paper links the public code repository directly. Only the public dataset access notice was inspected; gated files were not accessed. Publication metadata follows the final ACL record.

Sources reviewed

Source review only; experiments were not rerun for this collection. Cite the original authors using the linked paper or BibTeX.

Primary sources

  1. https://aclanthology.org/2026.acl-long.1750/
  2. https://aclanthology.org/2026.acl-long.1750.pdf
  3. https://arxiv.org/abs/2607.04727
  4. https://arxiv.org/html/2607.04727v1
  5. https://github.com/nth2000/Dashboard2Code
  6. https://huggingface.co/datasets/nth2000/Dashboard2Code