Imagen de Portada

Impact of Pretraining Word Co-occurrence on Compositional Generalization in Multimodal Models

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Qu, Helen, Xie, Sang Michael
Formato:	Preprint
Publicado:	2025
Materias:	Computer Vision and Pattern Recognition Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Connect Later: Improving Fine-tuning for Robustness with Targeted Augmentations
por: Qu, Helen, et al.
Publicado: (2024)

Pretrain Where? Investigating How Pretraining Data Diversity Impacts Geospatial Foundation Model Performance
por: Kaur, Amandeep, et al.
Publicado: (2026)

LMFusion: Adapting Pretrained Language Models for Multimodal Generation
por: Shi, Weijia, et al.
Publicado: (2024)

Context-Aware Multimodal Pretraining
por: Roth, Karsten, et al.
Publicado: (2024)

Towards Knowledge Guided Pretraining Approaches for Multimodal Foundation Models: Applications in Remote Sensing
por: Ravirathinam, Praveen, et al.
Publicado: (2024)

Foundation Model-oriented Robustness: Robust Image Model Evaluation with Pretrained Models
por: Zhang, Peiyan, et al.
Publicado: (2023)

A Practitioner's Guide to Continual Multimodal Pretraining
por: Roth, Karsten, et al.
Publicado: (2024)

Draw Like an Artist: Complex Scene Generation with Diffusion Model via Composition, Painting, and Retouching
por: Liu, Minghao, et al.
Publicado: (2024)

Pretrained Visual Uncertainties
por: Kirchhof, Michael, et al.
Publicado: (2024)

SpectralEarth-FM: Bringing Hyperspectral Imagery into Multimodal Earth Observation Pretraining
por: Braham, Nassim Ait Ali, et al.
Publicado: (2026)

Simplifying Knowledge Transfer in Pretrained Models
por: Jain, Siddharth, et al.
Publicado: (2025)

COMPRER: A Multimodal Multi-Objective Pretraining Framework for Enhanced Medical Image Representation
por: Lutsker, Guy, et al.
Publicado: (2024)

Is Large-Scale Pretraining the Secret to Good Domain Generalization?
por: Teterwak, Piotr, et al.
Publicado: (2024)

Why Fine-grained Labels in Pretraining Benefit Generalization?
por: Hong, Guan Zhe, et al.
Publicado: (2024)

SILMM: Self-Improving Large Multimodal Models for Compositional Text-to-Image Generation
por: Qu, Leigang, et al.
Publicado: (2024)

No "Zero-Shot" Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance
por: Udandarao, Vishaal, et al.
Publicado: (2024)

Pretrained Diffusion Models Are Inherently Skipped-Step Samplers
por: Xu, Wenju
Publicado: (2025)

Pretrained Image-Text Models are Secretly Video Captioners
por: Zhang, Chunhui, et al.
Publicado: (2025)

Progressive Compositionality in Text-to-Image Generative Models
por: Han, Evans Xu, et al.
Publicado: (2024)

Fantastic Gains and Where to Find Them: On the Existence and Prospect of General Knowledge Transfer between Any Pretrained Model
por: Roth, Karsten, et al.
Publicado: (2023)

2.5 Years in Class: A Multimodal Textbook for Vision-Language Pretraining
por: Zhang, Wenqi, et al.
Publicado: (2025)

Reflecting on the State of Rehearsal-free Continual Learning with Pretrained Models
por: Thede, Lukas, et al.
Publicado: (2024)

ROSE: Revolutionizing Open-Set Dense Segmentation with Patch-Wise Perceptual Large Multimodal Model
por: Han, Kunyang, et al.
Publicado: (2024)

Negative Label Guided OOD Detection with Pretrained Vision-Language Models
por: Jiang, Xue, et al.
Publicado: (2024)

Image Clustering via the Principle of Rate Reduction in the Age of Pretrained Models
por: Chu, Tianzhe, et al.
Publicado: (2023)

Accelerating Augmentation Invariance Pretraining
por: Lin, Jinhong, et al.
Publicado: (2024)

What Variables Affect Out-of-Distribution Generalization in Pretrained Models?
por: Harun, Md Yousuf, et al.
Publicado: (2024)

Learning Multimodal Latent Generative Models with Energy-Based Prior
por: Yuan, Shiyu, et al.
Publicado: (2024)

Enhancing Compositional Generalization via Compositional Feature Alignment
por: Wang, Haoxiang, et al.
Publicado: (2024)

Words That Make Language Models Perceive
por: Wang, Sophie L., et al.
Publicado: (2025)

Breaking the Likelihood-Quality Trade-off in Diffusion Models by Merging Pretrained Experts
por: Esfandiari, Yasin, et al.
Publicado: (2025)

Compositional Generalization Requires Linear, Orthogonal Representations in Vision Embedding Models
por: Uselis, Arnas, et al.
Publicado: (2026)

Medical Semantic Segmentation with Diffusion Pretrain
por: Li, David, et al.
Publicado: (2025)

Polyp Segmentation Generalisability of Pretrained Backbones
por: Sanderson, Edward, et al.
Publicado: (2024)

Multimodal Generalized Category Discovery
por: Su, Yuchang, et al.
Publicado: (2024)

Rethinking Generative Image Pretraining: How Far Are We From Scaling Up Next-Pixel Prediction?
por: Yan, Xinchen, et al.
Publicado: (2025)

Contrastive Test-Time Composition of Multiple LoRA Models for Image Generation
por: Meral, Tuna Han Salih, et al.
Publicado: (2024)

Self-Supervised Pretraining for Aerial Road Extraction
por: Polley, Rupert, et al.
Publicado: (2025)

Object-level Self-Distillation for Vision Pretraining
por: Hızlı, Çağlar, et al.
Publicado: (2025)

STARFlow2: Bridging Language Models and Normalizing Flows for Unified Multimodal Generation
por: Shen, Ying, et al.
Publicado: (2026)