Introduces KathaTrace protocol and KathaBench-25K benchmark to quantify Semantic Trajectory Gap (STG) as the loss of transition meaning in visualized narratives, reporting STG of 23.5 +/- 1.3 across generators.
arXiv preprint arXiv:2512.16853 , year=
16 Pith papers cite this work. Polarity classification is still indexing.
16
Pith papers citing it
citation-role summary
dataset 3
citation-polarity summary
years
2026 16roles
dataset 3polarities
use dataset 3representative citing papers
Flow-DPPO replaces PPO ratio clipping with an asymmetric KL divergence mask for flow models, claiming higher rewards, reduced forgetting, and stable multi-epoch training.
LENS shapes low-frequency eigen noise with a lightweight network to enable efficient, high-quality sampling in distilled diffusion models.
COMFYCLAW introduces skill evolution via graph editing, automatic reversion, VLM verification, and distillation of runs into reusable Agent Skills, achieving higher average scores than a verifier-only baseline across benchmarks.
NanoGen unifies DiT training on ImageNet and T2I, reveals negative Pearson correlations (-0.377 to -0.580) in method rankings across metrics from 21 models, and motivates DiffusionBench for holistic evaluation.
Compositionality emerges in neural networks only in a narrow depth-connectivity regime, with gradient descent converging to fractured solutions outside it.
RTDMD unifies KL minimization to a reward-tilted teacher into distribution matching plus reward terms, using AC-DMD in stage one and hybrid GRPO-style gradients plus SubGRPO in stage two to reach new SOTA on preference, aesthetic, and compositional metrics with 4-step generation on SD3, SD3.5, and F
EPIC introduces predicate-guided inference-time search that lifts compositional T2I prompt accuracy from 34% to 71% on GenEval2 with 31-81% lower execution costs.
Flow-Direct constructs a reusable non-parametric guidance field from the log-density ratio of base and target distributions using all accumulated reward samples for feedback-efficient guidance in flow models.
DynT2I-Eval creates fresh prompts via dimension decomposition and dynamic sampling to evaluate text-to-image models on text alignment, quality, and aesthetics while maintaining a stable leaderboard.
ELDiff integrates evidential learning into T2I diffusion via pixel evidence loss and token conflict loss to improve object-wise semantic consistency.
PaintBench provides a scalable deterministic benchmark for precise visual editing operations, revealing that even the best of 11 models achieves only 17.1% mIoU and that scores correlate strongly with applied data visualization editing performance.
BiDPO extends Diffusion DPO to bimodal preferences and adds region-aware guidance, improving compositional fidelity in text-to-image generation over prior methods.
FAGER is a new agentic framework that creates structured factual rubrics to evaluate and refine text-to-image outputs for implicit factual correctness across science, history, products, and culture.
Using understanding tasks as direct supervision during post-training improves image generation and editing in unified multimodal models.
Omni is a multimodal model whose native training on diverse data types enables context unrolling, allowing explicit reasoning across modalities to better approximate shared knowledge and improve downstream performance.
citing papers explorer
-
KathaTrace: Diagnosing Semantic Trajectory Collapse in Generated Visual Narratives
Introduces KathaTrace protocol and KathaBench-25K benchmark to quantify Semantic Trajectory Gap (STG) as the loss of transition meaning in visualized narratives, reporting STG of 23.5 +/- 1.3 across generators.
-
Flow-DPPO: Divergence Proximal Policy Optimization for Flow Matching Models
Flow-DPPO replaces PPO ratio clipping with an asymmetric KL divergence mask for flow models, claiming higher rewards, reduced forgetting, and stable multi-epoch training.
-
LENS: Low-Frequency Eigen Noise Shaping for Efficient Diffusion Sampling
LENS shapes low-frequency eigen noise with a lightweight network to enable efficient, high-quality sampling in distilled diffusion models.
-
COMFYCLAW: Self-Evolving Skill Harnesses for Image Generation Workflows
COMFYCLAW introduces skill evolution via graph editing, automatic reversion, VLM verification, and distillation of runs into reusable Agent Skills, achieving higher average scores than a verifier-only baseline across benchmarks.
-
DiffusionBench: On Holistic Evaluation of Diffusion Transformers
NanoGen unifies DiT training on ImageNet and T2I, reveals negative Pearson correlations (-0.377 to -0.580) in method rankings across metrics from 21 models, and motivates DiffusionBench for holistic evaluation.
-
Compositionality Emerges in a Narrow Depth-Connectivity Regime: Architecture Constraints and Solution Manifolds
Compositionality emerges in neural networks only in a narrow depth-connectivity regime, with gradient descent converging to fractured solutions outside it.
-
Reinforcing Few-step Generators via Reward-Tilted Distribution Matching
RTDMD unifies KL minimization to a reward-tilted teacher into distribution matching plus reward terms, using AC-DMD in stage one and hybrid GRPO-style gradients plus SubGRPO in stage two to reach new SOTA on preference, aesthetic, and compositional metrics with 4-step generation on SD3, SD3.5, and F
-
EPIC: Efficient Predicate-Guided Inference-Time Control for Compositional Text-to-Image Generation
EPIC introduces predicate-guided inference-time search that lifts compositional T2I prompt accuracy from 34% to 71% on GenEval2 with 31-81% lower execution costs.
-
Flow-Direct: Feedback-Efficient and Reusable Guidance for Flow Models via Non-Parametric Guidance Field
Flow-Direct constructs a reusable non-parametric guidance field from the log-density ratio of base and target distributions using all accumulated reward samples for feedback-efficient guidance in flow models.
-
DynT2I-Eval: A Dynamic Evaluation Framework for Text-to-Image Models
DynT2I-Eval creates fresh prompts via dimension decomposition and dynamic sampling to evaluate text-to-image models on text alignment, quality, and aesthetics while maintaining a stable leaderboard.
-
ELDiff: When Evidential Learning Meets Text-to-Image Diffusion
ELDiff integrates evidential learning into T2I diffusion via pixel evidence loss and token conflict loss to improve object-wise semantic consistency.
-
PaintBench: Deterministic Evaluation of Precise Visual Editing
PaintBench provides a scalable deterministic benchmark for precise visual editing operations, revealing that even the best of 11 models achieves only 17.1% mIoU and that scores correlate strongly with applied data visualization editing performance.
-
Compositional Text-to-Image Generation Via Region-aware Bimodal Direct Preference Optimization
BiDPO extends Diffusion DPO to bimodal preferences and adds region-aware guidance, improving compositional fidelity in text-to-image generation over prior methods.
-
FAGER: Factually Grounded Evaluation and Refinement of Text-to-Image Models
FAGER is a new agentic framework that creates structured factual rubrics to evaluate and refine text-to-image outputs for implicit factual correctness across science, history, products, and culture.
-
Steering Visual Generation in Unified Multimodal Models with Understanding Supervision
Using understanding tasks as direct supervision during post-training improves image generation and editing in unified multimodal models.
-
Context Unrolling in Omni Models
Omni is a multimodal model whose native training on diverse data types enables context unrolling, allowing explicit reasoning across modalities to better approximate shared knowledge and improve downstream performance.