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.
arXiv:2610.07689v1 Announce Type: new Abstract: Vision-Language Models (VLMs) such as CLIP excel in global semantic alignment but often lack fine-grained perceptual capabilities. This hinders dense p...
Visual Information-Guided Parallel Decoding for Diffusion Multimodal Large Language Models introduces the VIG‑Sampler, a method that prioritizes tokens for decoding based on their attention to image tokens and penalizes redundancy in image‑attention distributions. The approach aims to improve the quality of multimodal generation by selecting more informative tokens during diffusion decoding. Experiments on seven captioning and VQA benchmarks with three open‑source dMLLMs show that VIG‑Sampler outperforms the Info‑Gain Sampler by an average of 19.3 CIDEr points and achieves better COCO Caption results using only half as many decoding steps.
VISTA is a gradient‑based test‑time alignment framework designed for next‑scale visual autoregressive (VAR) image generation. It optimizes intermediate representations within the frozen transformer to enforce compositional constraints, without altering model weights or requiring extra training. Experiments on two benchmarks and two model scales show that VISTA improves compositional accuracy by up to 20% on a 2B backbone and 6% on an 8B backbone, while preserving image quality and enabling a smaller model to outperform a larger one.
Visual autoregressive (VAR) models have emerged as a fast, high-quality alternative to diffusion for text-to-image generation, but like diffusion models they exhibit persistent compositional failures,...
The paper introduces a 3D-CLIP encoder trained with structured hard negatives to improve vision‑language alignment for text‑to‑CT generation. This encoder drives a latent diffusion model that operates directly in 3D latent space, eliminating spatial artifacts from super‑resolution pipelines. Experiments on the CT‑RATE dataset show state‑of‑the‑art image fidelity and factual correctness across 18 pathological conditions, with lower inference time and GPU memory usage than competing methods.
arXiv:2610.02718v1 Announce Type: cross Abstract: Vision-language models such as CLIP exhibit strong semantic generalization, but remain limited in fine-grained visual perception. A recent work named...
arXiv:2608. 19871v1 Announce Type: new Abstract: Compositional Zero-Shot Learning (CZSL) aims to recognize unseen attribute-object compositions by leveraging knowledge of primitive concepts learned from seen compositions.
arXiv:2605. 13974v2 Announce Type: replace-cross Abstract: Diffusion Transformers (DiTs) and related flow-based architectures are now among the strongest text-to-image generators, yet the internal mechanisms through which prompts shape image semantics remain poorly understood.
arXiv:2602. 21397v2 Announce Type: replace-cross Abstract: Prompt learning has become a dominant paradigm for adapting vision-language models (VLMs) such as CLIP to downstream tasks without modifying pretrained weights.
arXiv:2605. 08245v4 Announce Type: replace-cross Abstract: Vision-Language Models (VLMs) increasingly power high-stakes applications, from medical imaging to autonomous systems, yet they routinely hallucinate, confidently describing content not present in the input.
D‑Scope is a framework that links the interpretation of sparse autoencoder (SAE) features in diffusion transformers (DiTs) to controllable image generation. It aggregates SigLIP‑2 embeddings of highly activating image patches into visual centroids, matches target text descriptions against these centroids, and retrieves individual features without per‑feature text annotations. The method provides visual evidence for each selection and uses spatially masked interventions to test decoder directions under fixed generation conditions, evaluated across 150 SAEs and a benchmark of 100 target concepts.
arXiv:2605. 25820v2 Announce Type: replace Abstract: Diffusion-based multimodal large language models (dMLLMs) decode by iteratively predicting tokens at multiple masked positions in parallel.