mlsanityData to 14 Oct 2026

SVG-Score: Human-Aligned Evaluation of Text-to-SVG Generation

arXiv 2609.03806 ↗First posted 3 Sep 2026cs.AI cs.CVcited by 0

On the map: LLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment

Abstract (arXiv)

Scalable Vector Graphics (SVG) generation is attracting increasing attention as generative models improve in expressiveness and controllability. Progress, however, is held back by the lack of domain-specific evaluation protocols: current practice relies on metrics designed for natural images, most notably CLIPScore, which was never trained on vector graphics and aligns only partially with human judgment. We introduce ours, a human-aligned evaluation framework for text-to-SVG generation. Through controlled caption and image perturbations, we first show that CLIP-based scores barely react to the errors SVG generators actually make, such as wrong colors, counts, and spatial relations, and that off-the-shelf Vision-Language Model (VLM) judges, while more sensitive, respond unevenly across error types and SVG styles. We then introduce a human-annotated dataset for textitSemantic Alignment, measuring how faithfully a generated SVG reflects its caption. Building on it, we develop two complementary evaluators: CLIP scorers adapted to vector graphics and then aligned to human preferences, for fast large-scale evaluation, and a VLM judge trained with supervised fine-tuning and reward-shaped reinforcement learning, for more expressive and interpretable assessment. Using both, we benchmark major open-source, commercial, and optimization-based SVG generators on an independent caption set.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Diffusion and multimodal model alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 59 of its 68 references that are in the map. First posted on arXiv 3 Sep 2026.

Builds on

59 of 68 references.

Closest ancestors · 2

Same direction, newest first

  1. 2026IntroSVG: Learning from Rendering Feedback for Text-to-SVG Generation via an Introspective Generator-Critic Framework10 Mar
  2. 2024VL-RewardBench: A Challenging Benchmark for Vision-Language Generative Reward Models26 Nov

Other references · 57

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  4. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  5. 2025Qwen3 Technical Reportcited by 8,820
  6. 2024GPT-4o System Cardcited by 4,588
  7. 2023Improved Baselines with Visual Instruction Tuningcited by 4,418
  8. 2022DreamFusion: Text-to-3D using 2D Diffusioncited by 3,025
  9. 2025Qwen3-VL Technical Reportcited by 2,793
  10. 2017Billion-scale similarity search with GPUscited by 2,497
  11. 2021CLIPScore: A Reference-free Evaluation Metric for Image Captioningcited by 2,424
  12. 2025InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Modelscited by 1,834
38 more
  1. 2025Gemma 3 Technical Reportcited by 1,824
  2. 2023ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generationcited by 1,045
  3. 2025OpenAI GPT-5 System Cardcited by 989
  4. 2023Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generationcited by 977
  5. 2023Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesiscited by 890
  6. 2024Autoregressive Model Beats Diffusion: Llama for Scalable Image Generationcited by 777
  7. 2024MLLM-as-a-Judge: Assessing Multimodal LLM-as-a-Judge with Vision-Language Benchmarkcited by 372
  8. 2023VIEScore: Towards Explainable Metrics for Conditional Image Synthesis Evaluationcited by 201
  9. 2021CLIPDraw: Exploring Text-to-Drawing Synthesis through Language-Image Encoderscited by 187
  10. 2020DeepSVG: A Hierarchical Generative Network for Vector Graphics Animationcited by 159
  11. 2024Prometheus-Vision: Vision-Language Model as a Judge for Fine-Grained Evaluationcited by 125
  12. 2021Im2Vec: Synthesizing Vector Graphics without Vector Supervisioncited by 116
  13. 2022VectorFusion: Text-to-SVG by Abstracting Pixel-Based Diffusion Modelscited by 112
  14. 2023StarVector: Generating Scalable Vector Graphics Code from Images and Textcited by 102
  15. 2019A Learned Representation for Scalable Vector Graphicscited by 91
  16. 2022Towards Layer-wise Image Vectorizationcited by 79
  17. 2023SVGDreamer: Text Guided SVG Generation with Diffusion Modelcited by 76
  18. 2023IconShop: Text-Guided Vector Icon Synthesis with Autoregressive Transformerscited by 75
  19. 2025OmniSVG: A Unified Scalable Vector Graphics Generation Modelcited by 74
  20. 2024Empowering LLMs to Understand and Generate Complex Vector Graphicscited by 66
  21. 2023DiffSketcher: Text Guided Vector Sketch Synthesis through Latent Diffusion Modelscited by 54
  22. 2025LayerTracer: Cognitive-Aligned Layered SVG Synthesis via Diffusion Transformercited by 43
  23. 2025Rendering-Aware Reinforcement Learning for Vector Graphics Generationcited by 39
  24. 2024Chat2SVG: Vector Graphics Generation with Large Language Models and Image Diffusion Modelscited by 38
  25. 2023DeepVecFont-v2: Exploiting Transformers to Synthesize Vector Fonts with Higher Qualitycited by 27
  26. 2024Score Distillation Sampling with Learned Manifold Correctivecited by 27
  27. 2024StrokeNUWA: Tokenizing Strokes for Vector Graphic Synthesiscited by 26
  28. 2024SVGEditBench: A Benchmark Dataset for Quantitative Assessment of LLM's SVG Editing Capabilitiescited by 26
  29. 2024NIVeL: Neural Implicit Vector Layers for Text-to-Vector Generationcited by 17
  30. 2025NeuralSVG: An Implicit Representation for Text-to-Vector Generationcited by 17
  31. 2023DualVector: Unsupervised Vector Font Synthesis with Dual-Part Representationcited by 16
  32. 2025SVGenius: Benchmarking LLMs in SVG Understanding, Editing and Generationcited by 16
  33. 2025UniSVG: A Unified Dataset for Vector Graphic Understanding and Generation with Multimodal Large Language Modelscited by 15
  34. 2025SVGen: Interpretable Vector Graphics Generation with Large Language Modelscited by 14
  35. 2025InternSVG: Towards Unified SVG Tasks with Multimodal Large Language Modelscited by 14
  36. 2025SVGEditBench V2: A Benchmark for Instruction-based SVG Editingcited by 11
  37. 2023Leveraging Large Language Models for Scalable Vector Graphics-Driven Image Understandingcited by 10
  38. 2024VectorPainter: Advanced Stylized Vector Graphics Synthesis Using Stroke-Style Priorscited by 7

50 most cited of 57 All 57 (JSON)

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2023Beyond Pixels: Exploring Human-Readable SVG Generation for Simple Images with Vision Language Models
  2. 2025SVGThinker: Instruction-Aligned and Reasoning-Driven Text-to-SVG Generation
  3. 2025DuetSVG: Unified Multimodal SVG Generation with Internal Visual Guidance
  4. 2024GenAI-Bench: Evaluating and Improving Compositional Text-to-Visual Generation
  5. 2025VectorEdits: A Dataset and Benchmark for Instruction-Based Editing of Vector Graphics
  6. 2024EvalMuse-40K: A Reliable and Fine-Grained Benchmark with Comprehensive Human Annotations for Text-to-Image Generation Model Evaluation
  7. 2026Compositional SVG Generation via VLM-Driven Hierarchical Semantic Parsing
  8. 2025VISTAR:A User-Centric and Role-Driven Benchmark for Text-to-Image Evaluation