Build error
Agents
24
VCode
šØ
Convert images to SVG code
Jinpeng Group
university
AI & ML interests
Multi-modality Model, Data-centric AI
Recent Activity
Papers
When the Prompt Becomes Visual: Vision-Centric Jailbreak Attacks for Large Image Editing Models
VCode: a Multimodal Coding Benchmark with SVG as Symbolic Visual Representation
Organization Card
š° [November 2025] We release VCode benchmark. See Website for details.
š° [October 2025] We release Chart2Code benchmark. See Website for details.
š° [Feb 2025]
We release TextAtlas5M, a dataset specifically designed for training and evaluating multimodal generation models on dense-text image generation.
datasets 14
CSU-JPG/Chart2Code
Viewer ā¢ Updated ā¢ 2.19k ā¢ 10.3k ā¢ 3
CSU-JPG/LabInstruct
Viewer ā¢ Updated ā¢ 3.93k ā¢ 80
CSU-JPG/MIND
Updated ā¢ 2.22k ā¢ 5
CSU-JPG/ChartJudgeBench
Viewer ā¢ Updated ā¢ 1.65k ā¢ 48
CSU-JPG/VVA-Bench
Updated ā¢ 288 ā¢ 1
CSU-JPG/IESBench
Updated ā¢ 705 ā¢ 8
CSU-JPG/Textground4M
Viewer ā¢ Updated ā¢ 4.1M ā¢ 7.33k ā¢ 2
CSU-JPG/VPBench
Viewer ā¢ Updated ā¢ 1.06k ā¢ 52 ā¢ 1
CSU-JPG/VisPrompt5M
Viewer ā¢ Updated ā¢ 5.36M ā¢ 2.04k ā¢ 9
CSU-JPG/TextEditBench
Viewer ā¢ Updated ā¢ 4 ā¢ 521 ā¢ 1