mlsanityData to 7 Oct 2026

VCG-Bench: Towards A Unified Visual-Centric Benchmark for Structured Generation and Editing

arXiv 2605.15677 ↗First posted 15 May 2026cs.CL cs.CVcited by 6

On the map: Multimodal LLMs / vision-language models → Chart understanding and anomaly detection

Abstract (arXiv)

Despite the rapid advancements in Vision-Language Models (VLMs), a critical gap remains in their ability to handle structured, controllable diagrammatic tasks essential for professional workflows. Existing methods predominantly rely on pixel-based synthesis, which operates in probabilistic pixel spaces and is inherently limited in editability and fidelity. Instead, we propose a new Diagram-as-Code paradigm with symbolic logic that leverages mxGraph Extensible Markup Language (XML) for precise diagram generation and editing. We present VCG-Bench, a unified benchmark for visual-centric textttmxGraph tasks. VCG-Bench comprises: (1) a taxonomized dataset of 1,449 diverse diagrams spanning 6 domains and 15 sub-domains, (2) a paradigm definition that integrates Generation (Vision-to-Code) and Editability (Code-to-Code), (3) a Tailored Evaluation Protocol employing multi-dimensional metrics such as textttmxGraph Execution Success Rate, Style Consistency Score (SCS), etc. Experimental results highlight the challenges faced by current State-of-the-Art (SOTA) VLMs in structured fidelity and instruction compliance, reflecting their vision and reasoning capabilities.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Placed in Chart understanding and anomaly detection, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 57 of its 73 references that are in the map. First posted on arXiv 15 May 2026.

Builds on

57 of 73 references.

Closest ancestors · 6

Same direction, newest first

  1. 2026RealChart2Code: Advancing Chart-to-Code Generation with Real Data and Multi-Task Evaluation26 Mar
  2. 2025SVGenius: Benchmarking LLMs in SVG Understanding, Editing and Generation3 Jun
  3. 2024FlowLearn: Evaluating Large Vision-Language Models on Flowchart Understanding6 Jul
  4. 2024CharXiv: Charting Gaps in Realistic Chart Understanding in Multimodal LLMs26 Jun
  5. 2024ChartMimic: Evaluating LMM's Cross-Modal Reasoning Capability via Chart-to-Code Generation14 Jun
  6. 2024Design2Code: Benchmarking Multimodal Code Generation for Automated Front-End Engineering5 Mar

Other references · 51

  1. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  4. 2023Adding Conditional Control to Text-to-Image Diffusion Modelscited by 5,466
  5. 2023Improved Baselines with Visual Instruction Tuningcited by 4,362
  6. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,290
  7. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,151
  8. 2022InstructPix2Pix: Learning to Follow Image Editing Instructionscited by 2,591
  9. 2015VQA: Visual Question Answeringcited by 2,489
  10. 2022Prompt-to-Prompt Image Editing with Cross Attention Controlcited by 2,409
  11. 2023Code Llama: Open Foundation Models for Codecited by 2,105
  12. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,044
38 more
  1. 2024Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scalingcited by 1,573
  2. 2022ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoningcited by 1,449
  3. 2023MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilitiescited by 1,221
  4. 2024DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligencecited by 1,154
  5. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,143
  6. 2025Qwen-Image Technical Reportcited by 973
  7. 2024Are We on the Right Way for Evaluating Large Vision-Language Models?cited by 893
  8. 2016A Diagram Is Worth A Dozen Imagescited by 793
  9. 2023StarCoder: may the source be with you!cited by 790
  10. 2025DeepSeek-V3.2: Pushing the Frontier of Open Large Language Modelscited by 755
  11. 2023MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image Editingcited by 615
  12. 2025Kimi K2: Open Agentic Intelligencecited by 523
  13. 2024MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmarkcited by 494
  14. 2024StarCoder 2 and The Stack v2: The Next Generationcited by 484
  15. 2017FigureQA: An Annotated Figure Dataset for Visual Reasoningcited by 314
  16. 2023LVLM-eHub: A Comprehensive Evaluation Benchmark for Large Vision-Language Modelscited by 217
  17. 2020DeepSVG: A Hierarchical Generative Network for Vector Graphics Animationcited by 158
  18. 2024InFoBench: Evaluating Instruction Following Ability in Large Language Modelscited by 128
  19. 2021Im2Vec: Synthesizing Vector Graphics without Vector Supervisioncited by 114
  20. 2017pix2code: Generating Code from a Graphical User Interface Screenshotcited by 113
  21. 2022VectorFusion: Text-to-SVG by Abstracting Pixel-Based Diffusion Modelscited by 112
  22. 2023IconShop: Text-Guided Vector Icon Synthesis with Autoregressive Transformerscited by 73
  23. 2025ChunkKV: Semantic-Preserving KV Cache Compression for Efficient Long-Context LLM Inferencecited by 57
  24. 2024MMCode: Benchmarking Multimodal Large Language Models for Code Generation with Visually Rich Programming Problemscited by 38
  25. 2024From Words to Structured Visuals: A Benchmark and Framework for Text-to-Diagram Generation and Editingcited by 29
  26. 2026STEP3-VL-10B Technical Reportcited by 28
  27. 2024LongGenBench: Long-context Generation Benchmarkcited by 27
  28. 2024HAIChart: Human and AI Paired Visualization Systemcited by 26
  29. 2024SVGEditBench: A Benchmark Dataset for Quantitative Assessment of LLM's SVG Editing Capabilitiescited by 25
  30. 2019Sketch2Code: Transformation of Sketches to UI in Real-time Using Deep Neural Networkcited by 21
  31. 2025Factuality Matters: When Image Generation and Editing Meet Structured Visualscited by 16
  32. 2025VisJudge-Bench: Aesthetics and Quality Assessment of Visualizationscited by 16
  33. 2025Can Compressed LLMs Truly Act? An Empirical Evaluation of Agentic Capabilities in LLM Compressioncited by 15
  34. 2025VCode: a Multimodal Coding Benchmark with SVG as Symbolic Visual Representationcited by 15
  35. 2025Draw with Thought: Unleashing Multimodal Reasoning for Scientific Diagram Generationcited by 14
  36. 2025PlotCraft: Pushing the Limits of LLMs for Complex and Interactive Data Visualizationcited by 11
  37. 2025SVGEditBench V2: A Benchmark for Instruction-based SVG Editingcited by 9
  38. 2025Flow2Code: Evaluating Large Language Models for Flowchart-based Code Generation Capabilitycited by 7

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 6.

6 papers.

  1. 2026DataMagic: Transforming Tabular Data into Data Insight Videocited by 2
  2. 2026DrawAI: Agentic Benchmark and Workflow for Making Raster Images Editablecited by 1
  3. 2026DataMagic: Authoring Data Videos through Declarative Multi-Agent Orchestrationcited by 1
  4. 2026Edit2TikZ: A Comprehensive and Challenging Benchmark for Scientific Figure Editing with TikZcited by 0
  5. 2026ReFigBench: Benchmarking Scientific Figure Reconstruction as Editable PowerPoint Artifactscited by 0
  6. 2026InfoEdit: Probing Global Layout Reasoning in Infographic Editingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025PlotGen-Bench: Evaluating VLMs on Generating Visualization Code from Diverse Plots across Multiple Libraries
  2. 2026VisEditBench: Can Vision-Language Models Edit Visualization Code from Multimodal Feedback?
  3. 2026GenAI-DrawIO-Creator: A Framework for Automated Diagram Generation
  4. 2025See it. Say it. Sorted: Agentic System for Compositional Diagram Generation
  5. 2025Flowchart2Mermaid: A Vision-Language Model Powered System for Converting Flowcharts into Editable Diagram Code
  6. 2025OpusAnimation: Code-Based Dynamic Chart Generation
  7. 2025StarFlow: Generating Structured Workflow Outputs From Sketch Images
  8. 2026Render-in-the-Loop: Vector Graphics Generation via Visual Self-Feedback