Cover Image

When Worse is Better: Navigating the compression-generation tradeoff in visual tokenization

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ramanujan, Vivek, Tirumala, Kushal, Aghajanyan, Armen, Zettlemoyer, Luke, Farhadi, Ali
Format:	Preprint
Published:	2024
Subjects:	Computer Vision and Pattern Recognition Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

VideoNet: A Large-Scale Dataset for Domain-Specific Action Recognition
by: Yadav, Tanush, et al.
Published: (2026)

For Better or For Worse? Learning Minimum Variance Features With Label Augmentation
by: Chidambaram, Muthu, et al.
Published: (2024)

From an Image to a Scene: Learning to Imagine the World from a Million 360 Videos
by: Wallingford, Matthew, et al.
Published: (2024)

Matryoshka Representation Learning
by: Kusupati, Aditya, et al.
Published: (2022)

CAT: Content-Adaptive Image Tokenization
by: Shen, Junhong, et al.
Published: (2025)

Contrastive Flow Matching
by: Stoica, George, et al.
Published: (2025)

Few-shot multi-token DreamBooth with LoRa for style-consistent character generation
by: Pascual, Ruben, et al.
Published: (2025)

LookupViT: Compressing visual information to a limited number of tokens
by: Koner, Rajat, et al.
Published: (2024)

Comparison of Autoencoders for tokenization of ASL datasets
by: Praun-Petrovic, Vouk, et al.
Published: (2025)

Bayesian computation with generative diffusion models by Multilevel Monte Carlo
by: Haji-Ali, Abdul-Lateef, et al.
Published: (2024)

Reconstruction Alignment Improves Unified Multimodal Models
by: Xie, Ji, et al.
Published: (2025)

Using Machine Learning for move sequence visualization and generation in climbing
by: Rimbot, Thomas, et al.
Published: (2025)

Reducing catastrophic forgetting of incremental learning in the absence of rehearsal memory with task-specific token
by: Choi, Young Jo, et al.
Published: (2024)

SCE-LITE-HQ: Smooth visual counterfactual explanations with generative foundation models
by: Zeid, Ahmed, et al.
Published: (2026)

Register and [CLS] tokens yield a decoupling of local and global features in large ViTs
by: Lappe, Alexander, et al.
Published: (2025)

Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion
by: Chen, Boyuan, et al.
Published: (2024)

LookWhen? Fast Video Recognition by Learning When, Where, and What to Compute
by: Salamatian, Ali, et al.
Published: (2026)

This Looks Better than That: Better Interpretable Models with ProtoPNeXt
by: Willard, Frank, et al.
Published: (2024)

Mixture-of-Mamba: Enhancing Multi-Modal State-Space Models with Modality-Aware Sparsity
by: Liang, Weixin, et al.
Published: (2025)

Posterior Augmented Flow Matching
by: Stoica, George, et al.
Published: (2026)

Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model
by: Zhou, Chunting, et al.
Published: (2024)

Text Quality-Based Pruning for Efficient Training of Language Models
by: Sharma, Vasu, et al.
Published: (2024)

tenSVD algorithm for compression
by: Gallo, Michele
Published: (2025)

ViTok-v2: Scaling Native Resolution Auto-Encoders to 5 Billion Parameters
by: Hansen-Estruch, Philippe, et al.
Published: (2026)

Sample what you cant compress
by: Birodkar, Vighnesh, et al.
Published: (2024)

CycleGAN with Better Cycles
by: Wang, Tongzhou, et al.
Published: (2024)

Imitating What Works: Simulation-Filtered Modular Policy Learning from Human Videos
by: Zhai, Albert J., et al.
Published: (2026)

Variational Bayes image restoration with compressive autoencoders
by: Biquard, Maud, et al.
Published: (2023)

Improving MoE Compute Efficiency by Composing Weight and Data Sparsity
by: Kilian, Maciej, et al.
Published: (2026)

Towards Better Understanding Attribution Methods
by: Rao, Sukrut, et al.
Published: (2022)

Unpaired Translation of Point Clouds for Modeling Detector Response
by: Li, Mingyang, et al.
Published: (2025)

Improving Visual Grounding by Encouraging Consistent Gradient-based Explanations
by: Yang, Ziyan, et al.
Published: (2022)

Robustly overfitting latents for flexible neural image compression
by: Perugachi-Diaz, Yura, et al.
Published: (2024)

Are Object-Centric Representations Better At Compositional Generalization?
by: Kapl, Ferdinand, et al.
Published: (2026)

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

Are Bias Mitigation Techniques for Deep Learning Effective?
by: Shrestha, Robik, et al.
Published: (2021)

Robustness to distribution shifts of compressed networks for edge devices
by: Shen, Lulan, et al.
Published: (2024)

Efficient training for compact compression models via sequential distillation
by: Rodrigues, Caroline Mazini, et al.
Published: (2026)

Disentangling Mean Embeddings for Better Diagnostics of Image Generators
by: Gruber, Sebastian G., et al.
Published: (2024)

Better Together: Evaluating the Complementarity of Earth Embedding Models
by: van der Plas, Thijs L, et al.
Published: (2026)