AI Paper Atlas

Research map / Video / Video-language understanding

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

Yang Chen, Minghao Liu, Yufan Shen, Yunwen Li, Tianyuan Huang, Xinyu Fang, Tianyu Zheng, Wenxuan Huang, Cheng Yang, Licheng Wen, Xuemeng Yang, Daocheng Fu and 8 more

ICLR 2026 · Video › Video-language understanding · approach: Event and anomaly reasoning in video · 16 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
enable thinking mode on Claude-Sonnet-4Claude-Sonnet-4 · IWR-BenchFinal Score: 34.00 -> 34.62Improved
enable thinking mode on Claude-Sonnet-4Claude-Sonnet-4 · IWR-BenchInteractive Functionality Score: 22.29 -> 23.65Improved
enable thinking mode on Claude-Opus-4Claude-Opus-4 · IWR-BenchFinal Score: 33.33 -> 34.13Improved
enable thinking mode on Claude-Opus-4Claude-Opus-4 · IWR-BenchInteractive Functionality Score: 21.83 -> 23.61Improved
enable thinking mode on Gemini-2.5-ProGemini-2.5-Pro · IWR-BenchFinal Score: 30.31 -> 30.36No clear effect
enable thinking mode on Gemini-2.5-ProGemini-2.5-Pro · IWR-BenchInteractive Functionality Score: 20.51 -> 21.65Improved
enable thinking mode on Gemini-2.5-ProGemini-2.5-Pro · IWR-BenchVisual Fidelity Score: 53.18 -> 50.67Hurt
move from static content (L1) to interactive tasks (L2 simple state manipulation)GPT-5 · IWR-BenchFinal Score: 61.85 -> 35.43Hurt
scale up from Qwen2.5-VL-7B to 72BQwen2.5-VL · IWR-BenchFinal Score: 15.47 -> 23.61Improved
use video-specialized model instead of general LVLM (VideoLLaMA3-7B vs Qwen2.5-VL-7B)7-8B open-source models · IWR-BenchFinal Score: 13.67 vs 15.47No clear effect
set final-score weighting alpha to 0.7 selected by human-alignment grid search (ablation)Final Score metric · IWR-Bench human alignmentSpearman correlation: rho=0.65Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy