# IWR-Bench: Can LVLMs reconstruct interactive webpage from a user interaction video?

> Tests reconstruction of interactive webpages from user-interaction videos and supplied static assets. Its 113 tasks replay annotated action sequences to assess visual fidelity and functional behavior across observed page states.

- Authors: Yang Chen; Minghao Liu; Yufan Shen; Yunwen Li; Tianyuan Huang; Xinyu Fang; Tianyu Zheng; Wenxuan Huang; Cheng Yang; Daocheng Fu; Jianbiao Mei; Rong Wu; Licheng Wen; Xuemeng Yang; Song Mao; Qunshu Lin; Zhi Yu; Yongliang Shen; Yu Qiao; Botian Shi
- Venue: ICLR 2026
- Status: Conference paper
- Publication / posting date: 2026 (exact date not verified)
- Date source: Publication year stated by the official ICLR proceedings; no finer publication date is asserted.
- First arXiv submission: September 29, 2025
- arXiv: 2509.24709
- Reviewed arXiv version: v3
- Sources reviewed: 2026-10-16
- Added to the catalog: 2026-10-16
- Original paper: <https://proceedings.iclr.cc/paper_files/paper/2026/hash/55bfedfd31489e5ae83c9ce8eec7b0e1-Abstract-Conference.html>
- PDF: <https://proceedings.iclr.cc/paper_files/paper/2026/file/55bfedfd31489e5ae83c9ce8eec7b0e1-Paper-Conference.pdf>
- Code: <https://github.com/SIGMME/IWR-Bench>
- Project: <https://sigmme.github.io/IWR-Bench/>
- BibTeX: <https://awesomerea.org/papers/citations/paper-iwr-bench.bib>
- Catalog page: <https://awesomerea.org/papers/paper-iwr-bench/>
- Record JSON: <https://awesomerea.org/papers/paper-iwr-bench/record.json>
- Chinese version: [简体中文](https://awesomerea.org/zh/papers/paper-iwr-bench/index.md)

## Scope and limitations

The required output is one HTML file; the paper explicitly instructs models to mock implied backend logic with static JavaScript data. Trajectory coverage and model-based judgments do not prove complete website or backend equivalence.

## Why it fits this collection

Provides a direct behavioral reconstruction benchmark beyond static screenshots while making the frontend/backend boundary explicit.

## Publication and review notes

Authors and order follow the published PDF: 20 authors. The proceedings landing page orders some middle authors differently; arXiv v3 and the project page list 21, additionally including Yunfei Zhao. The 2025 preprint date is separate from ICLR 2026 publication.

## Primary sources

1. <https://proceedings.iclr.cc/paper_files/paper/2026/hash/55bfedfd31489e5ae83c9ce8eec7b0e1-Abstract-Conference.html>
2. <https://proceedings.iclr.cc/paper_files/paper/2026/file/55bfedfd31489e5ae83c9ce8eec7b0e1-Paper-Conference.pdf>
3. <https://arxiv.org/abs/2509.24709>
4. <https://github.com/SIGMME/IWR-Bench>
5. <https://github.com/SIGMME/IWR-Bench/blob/891ca67cbd940961245d3eea883f211637e03490/README.md>
6. <https://sigmme.github.io/IWR-Bench/>
7. <https://huggingface.co/datasets/IWR-Bench/IWR-Bench>

## About this entry

Reviewed against the paper and its primary sources. Cite the original authors using the linked paper or BibTeX.
Original editorial content: MIT. Linked projects, papers and assets retain their own rights and licenses.