A joint latent diffusion model with cross-layer self-attention and disjoint sampling separates reflection and transmission layers from single images more effectively than prior methods on real-world benchmarks.
Test- time alignment of diffusion models without reward over- optimization
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
baseline 1
citation-polarity summary
years
2026 8verdicts
UNVERDICTED 8representative citing papers
IPR improves valid solution rates on MNIST Sudoku from 55.8% to 75.0% by iteratively refining partial regions in sequential diffusion models without external verifiers or reward models.
A new discrete diffusion model for scene graph generation from text captures object-relation dependencies via hierarchical constraints and training-free conditioning, yielding better graph metrics and downstream image alignment than prior baselines.
MSDDA derives a closed-form optimal reverse denoising distribution for multi-objective diffusion alignment that is exactly equivalent to step-level RL fine-tuning with no approximation error.
IMPFM is a multi-particle flow-map sampling method with sequential posterior sharing and interaction-aware correction that targets a KL-tilted distribution for global exploration in online feedback search.
SPARROW is a black-box optimization method that treats a fixed generative sampler as a structured proposal operator and applies rank-based selection over evaluated candidates to achieve low-budget optimization with asymptotic convergence guarantees over the sampler support.
AdaScope adaptively selects optimal RL intervention points during diffusion denoising by monitoring structural and semantic changes, delivering 66% higher performance at 59% lower cost than full-trajectory RL baselines.
VASR separates continuation and residual variance in reward-guided diffusion SMC, using optimal mass allocation and systematic resampling to achieve up to 26% better FID scores and faster runtimes than prior SMC and MCTS methods.
citing papers explorer
-
Reflection Separation from a Single Image via Joint Latent Diffusion
A joint latent diffusion model with cross-layer self-attention and disjoint sampling separates reflection and transmission layers from single images more effectively than prior methods on real-world benchmarks.
-
Inference-Time Scaling in Diffusion Models through Iterative Partial Refinement
IPR improves valid solution rates on MNIST Sudoku from 55.8% to 75.0% by iteratively refining partial regions in sequential diffusion models without external verifiers or reward models.
-
Dependency-Aware Discrete Diffusion for Scene Graph Generation
A new discrete diffusion model for scene graph generation from text captures object-relation dependencies via hierarchical constraints and training-free conditioning, yielding better graph metrics and downstream image alignment than prior baselines.
-
Step-level Denoising-time Diffusion Alignment with Multiple Objectives
MSDDA derives a closed-form optimal reverse denoising distribution for multi-objective diffusion alignment that is exactly equivalent to step-level RL fine-tuning with no approximation error.
-
Sequentially-Controlled Interactive Multi-Particle Flow-Maps for Online Feedback-Driven Search
IMPFM is a multi-particle flow-map sampling method with sequential posterior sharing and interaction-aware correction that targets a KL-tilted distribution for global exploration in online feedback search.
-
Generative Refinement for Low-Budget Black-Box Optimization
SPARROW is a black-box optimization method that treats a fixed generative sampler as a structured proposal operator and applies rank-based selection over evaluated candidates to achieve low-budget optimization with asymptotic convergence guarantees over the sampler support.
-
Do Less, Achieve More: Do We Need Every-Step Optimization for RL Fine-tuning of Diffusion Models?
AdaScope adaptively selects optimal RL intervention points during diffusion denoising by monitoring structural and semantic changes, delivering 66% higher performance at 59% lower cost than full-trajectory RL baselines.
-
VASR: Variance-Aware Systematic Resampling for Reward-Guided Diffusion
VASR separates continuation and residual variance in reward-guided diffusion SMC, using optimal mass allocation and systematic resampling to achieve up to 26% better FID scores and faster runtimes than prior SMC and MCTS methods.