196 citations · 716 across the 19 of their papers we have counts for
21 papers
PaliGemma: A versatile 3B VLM for transfer
Lucas Beyer, Andreas Steiner, André Susano Pinto +32
PaliGemma is an open Vision-Language Model (VLM) that is based on the SigLIP-So400m vision encoder and the Gemma-2B language model. It is trained to be a versatile and broadly know…
Routers in Vision Mixture of Experts: An Empirical Study
Tianlin Liu, Mathieu Blondel, Carlos Riquelme +1
Mixture-of-Experts (MoE) models are a promising way to scale up model capacity without significantly increasing computational cost. A key component of MoEs is the router, which dec…
From Sparse to Soft Mixtures of Experts
Joan Puigcerver, Carlos Riquelme, Basil Mustafa +1
Sparse mixture of expert architectures (MoEs) scale model capacity without significant increases in training or inference costs. Despite their success, MoEs suffer from a number of…
Patch n' Pack: NaViT, a Vision Transformer for any Aspect Ratio and Resolution
Mostafa Dehghani, Basil Mustafa, Josip Djolonga +12
The ubiquitous and demonstrably suboptimal choice of resizing images to a fixed resolution before processing them with computer vision models has not yet been successfully challeng…
Scaling Vision Transformers to 22 Billion Parameters
Mostafa Dehghani, Josip Djolonga, Basil Mustafa +39
The scaling of Transformers has driven breakthrough capabilities for language models. At present, the largest large language models (LLMs) contain upwards of 100B parameters. Visio…
Fast, Differentiable and Sparse Top-k: a Convex Analysis Perspective
Michael E. Sander, Joan Puigcerver, Josip Djolonga +2
The top-k operator returns a sparse vector, where the non-zero values correspond to the k largest values of the input. Unfortunately, because it is a discontinuous function, it is…
Sparse Upcycling: Training Mixture-of-Experts from Dense Checkpoints
Aran Komatsuzaki, Joan Puigcerver, James Lee-Thorp +6
Training large, deep neural networks to convergence can be prohibitively expensive. As a result, often only a small selection of popular, dense models are reused across different c…
On the Adversarial Robustness of Mixture of Experts
Joan Puigcerver, Rodolphe Jenatton, Carlos Riquelme +2
Adversarial robustness is a key desirable property of neural networks. It has been empirically shown to be affected by their sizes, with larger networks being typically more robust…
Sparsity-Constrained Optimal Transport
Tianlin Liu, Joan Puigcerver, Mathieu Blondel
Regularized optimal transport (OT) is now increasingly used as a loss or as a matching layer in neural networks. Entropy-regularized OT can be computed using the Sinkhorn algorithm…
PaLI: A Jointly-Scaled Multilingual Language-Image Model
Xi Chen, Xiao Wang, Soravit Changpinyo +26
Effective scaling and a flexible task interface enable large language models to excel at many tasks. We present PaLI (Pathways Language and Image model), a model that extends this…
Multimodal Contrastive Learning with LIMoE: the Language-Image Mixture of Experts
Basil Mustafa, Carlos Riquelme, Joan Puigcerver +2
Large sparsely-activated models have obtained excellent performance in multiple domains. However, such models are typically trained on a single modality at a time. We present the L…
Learning to Merge Tokens in Vision Transformers
Cedric Renggli, André Susano Pinto, Neil Houlsby +3
Transformers are widely applied to solve natural language understanding and computer vision tasks. While scaling up these architectures leads to improved performance, it often come…
Sparse MoEs meet Efficient Ensembles
James Urquhart Allingham, Florian Wenzel, Zelda E Mariet +10
Machine learning models based on the aggregated outputs of submodels, either at the activation or prediction levels, often exhibit strong performance compared to individual models.…
Scaling Vision with Sparse Mixture of Experts
Carlos Riquelme, Joan Puigcerver, Basil Mustafa +5
Sparsely-gated Mixture of Experts networks (MoEs) have demonstrated excellent scalability in Natural Language Processing. In Computer Vision, however, almost all performant network…
A Probabilistic Framework for Lexicon-based Keyword Spotting in Handwritten Text Images
E. Vidal, A. H. Toselli, J. Puigcerver
Query by String Keyword Spotting (KWS) is here considered as a key technology for indexing large collections of handwritten text images to allow fast textual access to the contents…
Deep Ensembles for Low-Data Transfer Learning
Basil Mustafa, Carlos Riquelme, Joan Puigcerver +3
In the low-data regime, it is difficult to train good supervised models from scratch. Instead practitioners turn to pre-trained models, leveraging transfer learning. Ensembling is…
Which Model to Transfer? Finding the Needle in the Growing Haystack
Cedric Renggli, André Susano Pinto, Luka Rimanic +4
Transfer learning has been recently popularized as a data-efficient alternative to training models from scratch, in particular for computer vision tasks where it provides a remarka…
Scalable Transfer Learning with Expert Models
Joan Puigcerver, Carlos Riquelme, Basil Mustafa +5
Transfer of pre-trained representations can improve sample efficiency and reduce computational requirements for new tasks. However, representations used for transfer are usually ge…
On Robustness and Transferability of Convolutional Neural Networks
Josip Djolonga, Jessica Yung, Michael Tschannen +11
Modern deep convolutional networks (CNNs) are often criticized for not generalizing under distributional shifts. However, several recent breakthroughs in transfer learning suggest…
Big Transfer (BiT): General Visual Representation Learning
Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai +4
Transfer of pre-trained representations improves sample efficiency and simplifies hyperparameter tuning when training deep neural networks for vision. We revisit the paradigm of pr…
A Large-scale Study of Representation Learning with the Visual Task Adaptation Benchmark
Xiaohua Zhai, Joan Puigcerver, Alexander Kolesnikov +14
Representation learning promises to unlock deep learning for the long tail of vision tasks without expensive labelled datasets. Yet, the absence of a unified evaluation for general…