arXiv Computer Vision October 5, 2026 By Fangzheng Wu, Brian Summa

Where to Look Is Not How to Fix: Pre-Denoising Diagnostics and Modality-Dependent Control in Diffusion Composition

Read the original on arXiv Computer Vision →

The paper introduces a controlled anchor–stress protocol to analyze compositional failures in text‑to‑image diffusion models. It presents a text‑only Compositional Stress Index (CSI) that distinguishes common from rare compositions across SD1.5, SDXL, and SD3, and conducts a six‑prompt localization study showing how different denoiser interventions affect outcomes such as residual minimization and color hit rate. The study reveals that diagnostic signals appear before denoising, but the intervention that exposes risk is not always the one that improves generation, highlighting modality‑ and architecture‑dependent responses.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computer Vision.

arXiv Machine Learning

Oct 8

ORCA: Hunting Compositional Failures in Text-to-Image Diffusion

ORCA (Orthogonal Residual Compositional Alignment) is a method that improves text-to-image diffusion models by aligning the diffusion transformer’s latent space with a low‑rank target derived from a frozen visual encoder. It introduces an auxiliary loss that uses a predictor with an orthogonal basis parameterised by a learned residual between T5 and CLIP embeddings, providing a prompt‑dependent signal for selecting the visual readout subspace. Experiments on three diffusion‑transformer backbones show that ORCA improves FID and GenEval scores, especially on attribute binding, spatial relations, and multi‑object prompts, without adding inference‑time cost.

By Arshia Hemmat, Amirhossein Vahidi, Amitis Shidani, Mohammad Vali Sanian, Hesam Asadollahzadeh, Aryan Yazdan Parast, Mohammad Lotfollahi

llms rag diffusion multimodal safety

Hugging Face Trending Papers

Jul 6

RADIANCE: Relative Adaptive Denoising with IP-Adapter for Novel Concept Enhancement

Text-to-image (T2I) diffusion models have achieved striking progress but still struggle to synthesize rare concepts involving unusual attribute-object pairings, often resulting in concept omission or semantic drift where a dominant entity overwhelms the generation. Tracing these failures to a lack of compositional balance during the denoising trajectory, we propose RADIANCE, a training-free framework that treats inference as a closed-loop feedback process.

diffusion efficiency benchmarks safety

arXiv AI

5d ago

DISTRACT: Degradation-Indexed Study of Textual Reference Across Competing Targets

arXiv:2610.28419v1 Announce Type: cross Abstract: Vision-language trackers (VLTs) accept descriptions intended to assist tracking, but accepting text does not establish that its content changes track...

By Dan Zlotnikov, Ohad Ben-Shahar

multimodal benchmarks

arXiv AI

Oct 1

Concept Unlearning via Cross-Attention Activation Projection for Diffusion Models

arXiv:2605.25765v2 Announce Type: replace-cross Abstract: Existing closed-form methods for concept unlearning in text-to-image diffusion models typically derive editing directions from fixed text emb...

By Saemi Moon, Suhyeon Jun, Seoyeon Lee, Dongwoo Kim

rag diffusion fine-tuning safety

arXiv Machine Learning

Jul 28

MaskAttn-SDXL: Controllable Region-Level Text-To-Image Generation

arXiv:2509. 15357v3 Announce Type: replace-cross Abstract: Diffusion models have achieved strong results in text-to-image generation, but important limitations remain as prompts become more structured and multi-object.

By Yu Chang, Jiahao Chen, Anzhe Cheng, Paul Bogdan

llms diffusion

arXiv Computer Vision

4d ago

How Do DiTs Denoise? An Aggregator-Predictor View of Attention and MLPs

The paper investigates how diffusion transformers (DiTs) denoise images by framing them as aggregator–predictor models that use context aggregation and patch prediction. It shows that for Gaussian Markov random field priors, a global Bayes denoiser can be approximated with a limited context, and that patchwise linear denoisers recover empirical local denoisers when aggregation is informative. Experiments confirm that DiTs trained on real images exhibit noise‑dependent denoising contexts and that replacing DiT MLPs with shared‑memory patch predictors yields competitive performance, while cross‑dataset component transplantation supports the proposed functional roles of attention and MLPs.

By Gyeonghun Kang, William Xingxu Chen, Xiang Cheng

llms diffusion fine-tuning