mlsanityData to 7 Oct 2026

YINYANG-ALIGN: Benchmarking Contradictory Objectives and Proposing Multi-Objective Optimization based DPO for Text-to-Image Alignment

arXiv 2502.03512 ↗First posted 5 Feb 2025cs.AIcited by 0

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

Abstract (arXiv)

Precise alignment in Text-to-Image (T2I) systems is crucial to ensure that generated visuals not only accurately encapsulate user intents but also conform to stringent ethical and aesthetic benchmarks. Incidents like the Google Gemini fiasco, where misaligned outputs triggered significant public backlash, underscore the critical need for robust alignment mechanisms. In contrast, Large Language Models (LLMs) have achieved notable success in alignment. Building on these advancements, researchers are eager to apply similar alignment techniques, such as Direct Preference Optimization (DPO), to T2I systems to enhance image generation fidelity and reliability. We present YinYangAlign, an advanced benchmarking framework that systematically quantifies the alignment fidelity of T2I systems, addressing six fundamental and inherently contradictory design objectives. Each pair represents fundamental tensions in image generation, such as balancing adherence to user prompts with creative modifications or maintaining diversity alongside visual coherence. YinYangAlign includes detailed axiom datasets featuring human prompts, aligned (chosen) responses, misaligned (rejected) AI-generated outputs, and explanations of the underlying contradictions.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Diffusion and multimodal model alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 31 of its 81 references that are in the map. First posted on arXiv 5 Feb 2025.

Builds on

31 of 81 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023Diffusion Model Alignment Using Direct Preference Optimization21 Nov
  2. 2023Aligning Text-to-Image Models using Human Feedback23 Feb

Other references · 29

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  4. 2023GPT-4 Technical Reportcited by 18,132
  5. 2022Training language models to follow instructions with human feedbackcited by 15,712
  6. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  7. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  8. 2023Visual Instruction Tuningcited by 7,792
  9. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 4,993
  10. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,290
  11. 2021Zero-Shot Text-to-Image Generationcited by 4,276
  12. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,634
17 more
  1. 2023Let's Verify Step by Stepcited by 3,598
  2. 2017Progressive Growing of GANs for Improved Quality, Stability, and Variationcited by 2,967
  3. 2017Decoupled Weight Decay Regularizationcited by 1,722
  4. 2024Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preferencecited by 1,151
  5. 2023AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedbackcited by 738
  6. 2021Carbon Emissions and Large Neural Network Trainingcited by 561
  7. 2020The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memescited by 485
  8. 2015A Neural Algorithm of Artistic Stylecited by 480
  9. 2023UltraFeedback: Boosting Language Models with Scaled AI Feedbackcited by 394
  10. 2017Wasserstein Auto-Encoderscited by 371
  11. 2020Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Modelscited by 168
  12. 2023HelpSteer: Multi-attribute Helpfulness Dataset for SteerLMcited by 140
  13. 2024LLaVA-Critic: Learning to Evaluate Multimodal Modelscited by 132
  14. 2024SAFREE: Training-Free and Adaptive Guard for Safe Text-to-Image And Video Generationcited by 125
  15. 2024Prometheus-Vision: Vision-Language Model as a Judge for Fine-Grained Evaluationcited by 123
  16. 2023What You See is What You Read? Improving Text-Image Alignment Evaluationcited by 105
  17. 2023RESI: An R Package for Robust Effect Sizescited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025REVEALER: Reinforcement-Guided Visual Reasoning for Element-Level Text-Image Alignment Evaluation
  2. 2025The Path to Reconciling Quality and Safety in Text-to-Image Generation: Dataset, Method, and Evaluation
  3. 2026TextAlign: Preference Alignment for Text Rendering with Hierarchical Rewards
  4. 2025Align Beyond Prompts: Evaluating World Knowledge Alignment in Text-to-Image Generation
  5. 2026Arena-T2I Hard: Benchmarking and Improving Faithfulness with Dependency-Aware Checklist
  6. 2025DETONATE: A Benchmark for Text-to-Image Alignment and Kernelized Direct Preference Optimization
  7. 2025TIT-Score: Evaluating Long-Prompt Based Text-to-Image Alignment via Text-to-Image-to-Text Consistency
  8. 2024Generalizing Alignment Paradigm of Text-to-Image Generation with Preferences through f-divergence Minimization