DiT-Reward converts pretrained DiT models into reward predictors that outperform HPSv3 on four benchmarks while providing 1.65x inference speedup.
hub Mixed citations
Rewarddance: Reward scaling in visual generation
Mixed citation behavior. Most common role is background (67%).
21
Pith papers citing it
Background
67% of classified citations
hub tools
citation-role summary
background 4
baseline 1
method 1
citation-polarity summary
representative citing papers
JudgeFit produces per-VLM physical video evaluation taxonomies that improve held-out accuracy by a mean 32% relative to a single global schema across 16 models from eight families.
DiffusionOPD applies online policy distillation from per-task teachers to a unified diffusion student, with a derived closed-form per-step KL objective that unifies SDE and ODE sampling via mean matching.
CaC presents a new spatiotemporal concentrating reward model for video anomalies, built on a novel large-scale dataset and three-stage training with RL and IoU rewards, claiming 25.7% accuracy gains and 11.7% anomaly reduction.
TMPO uses Softmax Trajectory Balance to match policy probabilities over multiple trajectories to a Boltzmann reward distribution, improving diversity by 9.1% in diffusion alignment tasks.
DiNa-LRM introduces a diffusion-native latent reward model using a noise-calibrated Thurstone likelihood on noisy states, matching VLM performance at lower compute in image alignment and preference optimization.
MixGRPO speeds up GRPO for flow-based image generators by restricting SDE sampling and optimization to a sliding window while using ODE elsewhere, cutting training time by up to 71% with better alignment performance.
A teacher-student reward model learns reasoning-conditioned score distributions for text-to-image images, yielding ~89% preference accuracy and a 41% net human-preference gain when used for generator optimization.
Finite-particle approximation of the Doob h-function causes reward hacking via two failure modes in reward-guided diffusion; a damping schedule corrects within-mode bias in Gaussian settings.
StressDream optimizes initial noise in diffusion video world models using VLM semantic and plausibility objectives to steer generations toward specified high-impact outcomes for improved policy evaluation.
DeScore decouples CoT reasoning from reward scoring in video reward models using a two-stage training process to improve generalization and avoid optimization bottlenecks of coupled generative RMs.
Edit-R1 builds a CoT-based reasoning reward model (RRM) via SFT and GCPO, then applies it with GRPO to improve image editing models such as FLUX.1-kontext.
Seedance 1.5 pro is a joint audio-visual generation model achieving high synchronization via dual-branch diffusion transformer and post-training optimizations.
Reward Forcing combines EMA-Sink tokens and Rewarded Distribution Matching Distillation to deliver state-of-the-art streaming video generation at 23.1 FPS without copying initial frames.
UniWorld-V2 applies policy optimization via DiffusionNFT and MLLM logit feedback with group filtering to reach state-of-the-art scores of 4.49 on ImgEdit and 7.83 on GEdit-Bench while remaining model-agnostic.
VRPO applies generative representation policy optimization to dynamically align diffusion features with pretrained visual encoders, claiming +1.8 FID gains and 2.3x faster training versus REPA.
Edit-GRPO decouples editing and preservation objectives via region-specific signals in a policy optimization framework to improve locality in image editing tasks.
A post-training pipeline for video generation models combines SFT, RLHF with novel GRPO, prompt enhancement, and inference optimization to improve visual quality, temporal coherence, and instruction following.
Seedance 2.0 is an updated multi-modal model for generating 4-15 second audio-video content at 480p/720p with support for up to 3 video, 9 image, and 3 audio references.
Seedream 4.0 unifies text-to-image synthesis, image editing, and multi-image composition in an efficient diffusion transformer pretrained on billions of pairs and accelerated to 1.8 seconds for 2K output.
citing papers explorer
-
DiT-Reward: Generative Representations for Text-to-Image Reward Modeling
DiT-Reward converts pretrained DiT models into reward predictors that outperform HPSv3 on four benchmarks while providing 1.65x inference speedup.
-
Each Judge Its Own Yardstick: Discovering Per-VLM Taxonomies for Physical Video Evaluation
JudgeFit produces per-VLM physical video evaluation taxonomies that improve held-out accuracy by a mean 32% relative to a single global schema across 16 models from eight families.
-
DiffusionOPD: A Unified Perspective of On-Policy Distillation in Diffusion Models
DiffusionOPD applies online policy distillation from per-task teachers to a unified diffusion student, with a derived closed-form per-step KL objective that unifies SDE and ODE sampling via mean matching.
-
CaC: Advancing Video Reward Models via Hierarchical Spatiotemporal Concentrating
CaC presents a new spatiotemporal concentrating reward model for video anomalies, built on a novel large-scale dataset and three-stage training with RL and IoU rewards, claiming 25.7% accuracy gains and 11.7% anomaly reduction.
-
TMPO: Trajectory Matching Policy Optimization for Diverse and Efficient Diffusion Alignment
TMPO uses Softmax Trajectory Balance to match policy probabilities over multiple trajectories to a Boltzmann reward distribution, improving diversity by 9.1% in diffusion alignment tasks.
-
Beyond VLM-Based Rewards: Diffusion-Native Latent Reward Modeling
DiNa-LRM introduces a diffusion-native latent reward model using a noise-calibrated Thurstone likelihood on noisy states, matching VLM performance at lower compute in image alignment and preference optimization.
-
MixGRPO: Unlocking Flow-based GRPO Efficiency with Mixed ODE-SDE
MixGRPO speeds up GRPO for flow-based image generators by restricting SDE sampling and optimization to a sliding window while using ODE elsewhere, cutting training time by up to 71% with better alignment performance.
-
Z-Reward: Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions
A teacher-student reward model learns reasoning-conditioned score distributions for text-to-image images, yielding ~89% preference accuracy and a 41% net human-preference gain when used for generator optimization.
-
Are we really tilting? The mechanics of reward guidance in flow and diffusion models
Finite-particle approximation of the Doob h-function causes reward hacking via two failure modes in reward-guided diffusion; a damping schedule corrects within-mode bias in Gaussian settings.
-
StressDream: Steering Video World Models for Robust Policy Evaluation and Improvement
StressDream optimizes initial noise in diffusion video world models using VLM semantic and plausibility objectives to steer generations toward specified high-impact outcomes for improved policy evaluation.
-
Think, then Score: Decoupled Reasoning and Scoring for Video Reward Modeling
DeScore decouples CoT reasoning from reward scoring in video reward models using a two-stage training process to improve generalization and avoid optimization bottlenecks of coupled generative RMs.
-
Leveraging Verifier-Based Reinforcement Learning in Image Editing
Edit-R1 builds a CoT-based reasoning reward model (RRM) via SFT and GCPO, then applies it with GRPO to improve image editing models such as FLUX.1-kontext.
-
Seedance 1.5 pro: A Native Audio-Visual Joint Generation Foundation Model
Seedance 1.5 pro is a joint audio-visual generation model achieving high synchronization via dual-branch diffusion transformer and post-training optimizations.
-
Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation
Reward Forcing combines EMA-Sink tokens and Rewarded Distribution Matching Distillation to deliver state-of-the-art streaming video generation at 23.1 FPS without copying initial frames.
-
Uniworld-V2: Reinforce Image Editing with Diffusion Negative-aware Finetuning and MLLM Implicit Feedback
UniWorld-V2 applies policy optimization via DiffusionNFT and MLLM logit feedback with group filtering to reach state-of-the-art scores of 4.49 on ImgEdit and 7.83 on GEdit-Bench while remaining model-agnostic.
-
Improving Visual Representation Alignment Generation with GRPO
VRPO applies generative representation policy optimization to dynamically align diffusion features with pretrained visual encoders, claiming +1.8 FID gains and 2.3x faster training versus REPA.
-
Edit-GRPO: A Locality-Preserving Policy Optimization Framework for Image Editing
Edit-GRPO decouples editing and preservation objectives via region-specific signals in a policy optimization framework to improve locality in image editing tasks.
-
A Systematic Post-Train Framework for Video Generation
A post-training pipeline for video generation models combines SFT, RLHF with novel GRPO, prompt enhancement, and inference optimization to improve visual quality, temporal coherence, and instruction following.
-
Seedance 2.0: Advancing Video Generation for World Complexity
Seedance 2.0 is an updated multi-modal model for generating 4-15 second audio-video content at 480p/720p with support for up to 3 video, 9 image, and 3 audio references.
-
Seedream 4.0: Toward Next-generation Multimodal Image Generation
Seedream 4.0 unifies text-to-image synthesis, image editing, and multi-image composition in an efficient diffusion transformer pretrained on billions of pairs and accelerated to 1.8 seconds for 2K output.
- NormGuard: Reward-Preserving Norm Constraints in Flow-Matching Reinforcement Learning