AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Pareto-Guided Optimal Transport for Multi-Reward Alignment

Ying Ba, Tianyu Zhang, Mohan Zhou, Yalong Bai, Wenyi Mo, Guiwei Zhang, Bing Su, Ji-Rong Wen

ICML 2026 · Image generation › Text-to-image diffusion · approach: Preference optimization for diffusion · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add optimal transport mapping to Pareto-bound targets (ablation)PG-OT · Parti-PromptsJDR4 (%): 10.78 -> 17.10Improved
add the online strategy on top of offline strategy (ablation)PG-OT · Parti-PromptsJDR2 (%): 38.54 -> 47.98Improved
use offline Pareto frontier only vs online only (ablation)PG-OT · Parti-PromptsJCR4 (%): 4.04 -> 7.23Hurt
use prompt-specific Pareto-bound targets with OT instead of weighted-sum multi-rewardSD3.5-Turbo + LoRA · Parti-PromptsJDR2 (%): 26.59 -> 47.98Improved
optimize a single reward (CLIP only) (ablation)SD3.5-Turbo + LoRA · CLIP-only optimizationHPS and HP change: HPS -2.78%, HP -4.38% at step 400Hurt
use prompt-specific bounds via weighted-sum without OT (ablation)SD3.5-Turbo + LoRA · Parti-PromptsJDR2 (%): 26.59 -> 29.84Mixed
use separate-constraint bounds without OT (ablation)SD3.5-Turbo + LoRA · Parti-PromptsHPS win rate (%): 82.97 -> 61.21Hurt
add weak reward models (CLIP, HPS) to joint optimizationmulti-reward weighted training · Parti-Promptsreward hackingHurt
use PG-OT vs single-reward HP in human studyPG-OT · DiffusionDBhuman win rate (%): 76.34 vs SD3.5-Turbo; 61.29 vs HPImproved
add VLM agent to remove weak rewardsPG-OT · Parti-PromptsJDR2Improved
vary entropy regularization strength epsilon (ablation)PG-OT · Appendix ErobustnessNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy