mlsanityData to 7 Oct 2026

MagicWand: A Universal Agent for Generation and Evaluation Aligned with User Preference

arXiv 2511.18352 ↗First posted 23 Nov 2025cs.CVcited by 1

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

Abstract (arXiv)

Recent advances in AIGC (Artificial Intelligence Generated Content) models have enabled significant progress in image and video generation. However, users still struggle to obtain content that aligns with their preferences due to the difficulty of crafting detailed prompts and the lack of mechanisms to retain their preferences. To address these challenges, we construct UniPrefer-100K, a large-scale dataset comprising images, videos, and associated text that describes the styles users tend to prefer. Based on UniPrefer-100K, we propose MagicWand, a universal generation and evaluation agent that enhances prompts based on user preferences, leverages advanced generation models for high-quality content, and applies preference-aligned evaluation and refinement. In addition, we introduce UniPreferBench, the first large-scale benchmark with over 120K annotations for assessing user preference alignment across diverse AIGC tasks. Experiments on UniPreferBench demonstrate that MagicWand consistently generates content and evaluations that are well aligned with user preferences across a wide range of scenarios.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Diffusion and multimodal model alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 35 of its 42 references that are in the map. First posted on arXiv 23 Nov 2025.

Builds on

35 of 42 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025EditScore: Unlocking Online RL for Image Editing via High-Fidelity Reward Modeling28 Sep

Other references · 34

  1. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  2. 2025Qwen3 Technical Reportcited by 8,195
  3. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,290
  4. 2025Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilitiescited by 3,655
  5. 2021CLIPScore: A Reference-free Evaluation Metric for Image Captioningcited by 2,388
  6. 2024CogVideoX: Text-to-Video Diffusion Models with An Expert Transformercited by 2,198
  7. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,906
  8. 2023VBench: Comprehensive Benchmark Suite for Video Generative Modelscited by 1,755
  9. 2024HunyuanVideo: A Systematic Framework For Large Video Generative Modelscited by 1,597
  10. 2023Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debatecited by 995
  11. 2025Qwen-Image Technical Reportcited by 973
  12. 2025FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Spacecited by 960
22 more
  1. 2022Tune-A-Video: One-Shot Tuning of Image Diffusion Models for Text-to-Video Generationcited by 958
  2. 2024Large Language Model based Multi-Agents: A Survey of Progress and Challengescited by 752
  3. 2023ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debatecited by 724
  4. 2022MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessmentcited by 514
  5. 2023FateZero: Fusing Attentions for Zero-shot Text-based Video Editingcited by 462
  6. 2024Understanding the planning of LLM agents: A surveycited by 414
  7. 2019Blind Image Quality Assessment Using A Deep Bilinear Convolutional Neural Networkcited by 301
  8. 2025OmniGen2: Towards Instruction-Aligned Multimodal Generationcited by 241
  9. 2025Seedream 4.0: Toward Next-generation Multimodal Image Generationcited by 233
  10. 2023TOPIQ: A Top-down Approach from Semantics to Distortions for Image Quality Assessmentcited by 220
  11. 2024Diffusion Model-Based Image Editing: A Surveycited by 212
  12. 2024VideoScore: Building Automatic Metrics to Simulate Fine-grained Human Feedback for Video Generationcited by 190
  13. 2017Video Generation From Textcited by 153
  14. 2023Consistent Video-to-Video Transfer Using Synthetic Datasetcited by 79
  15. 2025Scaling Instruction-Based Video Editing with a High-Quality Synthetic Datasetcited by 76
  16. 2025InsViE-1M: Effective Instruction-based Video Editing with Elaborate Dataset Constructioncited by 62
  17. 2025Pico-Banana-400K: A Large-Scale Dataset for Text-Guided Image Editingcited by 58
  18. 2025PromptEnhancer: A Simple Approach to Enhance Text-to-Image Models via Chain-of-Thought Prompt Rewritingcited by 38
  19. 2025LMM4LMM: Benchmarking and Evaluating Large-multimodal Image Generation with LMMscited by 24
  20. 2025LMM4Edit: Benchmarking and Evaluating Multimodal Image Editing with LMMscited by 22
  21. 2025LOVE: Benchmarking and Evaluating Text-to-Video Generation and Video-to-Text Interpretationcited by 10
  22. 2025TDVE-Assessor: Benchmarking and Evaluating the Quality of Text-Driven Video Editing with LMMscited by 10

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026EditRefiner: A Human-Aligned Agentic Framework for Image Editing Refinementcited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2024Multimodal Large Language Model is a Human-Aligned Annotator for Text-to-Image Generation
  2. 2024Preference Adaptive and Sequential Text-to-Image Generation
  3. 2024ViPer: Visual Personalization of Generative Models via Individual Preference Learning
  4. 2025Personalized Reward Modeling for Text-to-Image Generation
  5. 2023Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesis
  6. 2026Premier: Personalized Preference Modulation with Learnable User Embedding in Text-to-Image Generation
  7. 2025Instant Preference Alignment for Text-to-Image Diffusion Models
  8. 2026TailorMind: Towards Preference-Aligned Multimodal Content Generation