There are a thousand ways to caption an image. Contrastive Language Pretraining (CLIP) on the other hand, works by mapping an image and its caption to a single vector -- limiting how well CLIP-like models can represent the diverse ways to describe an image. In this work, we introduce Llip, Latent Language Image Pretraining, which models the diversity of captions that could match an image. Llip's vision encoder outputs a set of visual features that are mixed into a final representation by conditioning on information derived from the text. We show that Llip outperforms non-contextualized baselines like CLIP and SigLIP on a variety of tasks even with large-scale encoders. Llip improves zero-shot classification by an average of 2.9% zero-shot classification benchmarks with a ViT-G/14 encoder. Specifically, Llip attains a zero-shot top-1 accuracy of 83.5% on ImageNet outperforming a similarly sized CLIP by 1.4%. We also demonstrate improvement on zero-shot retrieval on MS-COCO by 6.0%. We provide a comprehensive analysis of the components introduced by the method and demonstrate that Llip leads to richer visual representations.

In this paper, we propose a novel model for a malware classification system based on Application Programming Interface (API) calls and opcodes, to improve classification accuracy. This system uses a novel design of combined Convolutional Neural Network and Long Short-Term Memory. We extract opcode sequences and API Calls from Windows malware samples for classification. We transform these features into N-grams (N = 2, 3, and 10)-gram sequences. Our experiments on a dataset of 9,749,57 samples produce high accuracy of 99.91% using the 8-gram sequences. Our method significantly improves the malware classification performance when using a wide range of recent deep learning architectures, leading to state-of-the-art performance. In particular, we experiment with ConvNeXt-T, ConvNeXt-S, RegNetY-4GF, RegNetY-8GF, RegNetY-12GF, EfficientNetV2, Sequencer2D-L, Swin-T, ViT-G/14, ViT-Ti, ViT-S, VIT-B, VIT-L, and MaxViT-B. Among these architectures, Swin-T and Sequencer2D-L architectures achieved high accuracies of 99.82% and 99.70%, respectively, comparable to our CNN-LSTM architecture although not surpassing it.

Scaling laws have been recently employed to derive compute-optimal model size (number of parameters) for a given compute duration. We advance and refine such methods to infer compute-optimal model shapes, such as width and depth, and successfully implement this in vision transformers. Our shape-optimized vision transformer, SoViT, achieves results competitive with models that exceed twice its size, despite being pre-trained with an equivalent amount of compute. For example, SoViT-400m/14 achieves 90.3% fine-tuning accuracy on ILSRCV2012, surpassing the much larger ViT-g/14 and approaching ViT-G/14 under identical settings, with also less than half the inference cost. We conduct a thorough evaluation across multiple tasks, such as image classification, captioning, VQA and zero-shot transfer, demonstrating the effectiveness of our model across a broad range of domains and identifying limitations. Overall, our findings challenge the prevailing approach of blindly scaling up vision models and pave a path for a more informed scaling.

Leveraging transfer learning has recently been shown to be an effective strategy for training large models with Differential Privacy (DP). Moreover, somewhat surprisingly, recent works have found that privately training just the last layer of a pre-trained model provides the best utility with DP. While past studies largely rely on algorithms like DP-SGD for training large models, in the specific case of privately learning from features, we observe that computational burden is low enough to allow for more sophisticated optimization schemes, including second-order methods. To that end, we systematically explore the effect of design parameters such as loss function and optimization algorithm. We find that, while commonly used logistic regression performs better than linear regression in the non-private setting, the situation is reversed in the private setting. We find that linear regression is much more effective than logistic regression from both privacy and computational aspects, especially at stricter epsilon values ($\epsilon < 1$). On the optimization side, we also explore using Newton's method, and find that second-order information is quite helpful even with privacy, although the benefit significantly diminishes with stricter privacy guarantees. While both methods use second-order information, least squares is effective at lower epsilons while Newton's method is effective at larger epsilon values. To combine the benefits of both, we propose a novel algorithm called DP-FC, which leverages feature covariance instead of the Hessian of the logistic regression loss and performs well across all $\epsilon$ values we tried. With this, we obtain new SOTA results on ImageNet-1k, CIFAR-100 and CIFAR-10 across all values of $\epsilon$ typically considered. Most remarkably, on ImageNet-1K, we obtain top-1 accuracy of 88\% under (8, $8 * 10^{-7}$)-DP and 84.3\% under (0.1, $8 * 10^{-7}$)-DP.

Google Brain researchers announced a two-billion-parameter deep-learning computer vision (CV) model. The model was trained on three billion pictures and obtained a new state-of-the-art record of 90.45 percent top-1 accuracy on ImageNet. The ViT-G/14 model is based on Google's latest Vision Transformers development (ViT). On numerous benchmarks, including ImageNet, ImageNet-v2, and VTAB-1k, ViT-G/14 beat prior state-of-the-art systems. For example, the accuracy gain on the few-shot picture identification challenge was more than five percentage points.

Researchers at Google Brain announced a deep-learning computer vision (CV) model containing two billion parameters. The model was trained on three billion images and achieved 90.45% top-1 accuracy on ImageNet, setting a new state-of-the-art record. The team described the model and experiments in a paper published on arXiv. The model, dubbed ViT-G/14, is based on Google's recent work on Vision Transformers (ViT). ViT-G/14 outperformed previous state-of-the-art solutions on several benchmarks, including ImageNet, ImageNet-v2, and VTAB-1k.

Attention-based neural networks such as the Vision Transformer (ViT) have recently attained state-of-the-art results on many computer vision benchmarks. Scale is a primary ingredient in attaining excellent results, therefore, understanding a model's scaling properties is a key to designing future generations effectively. While the laws for scaling Transformer language models have been studied, it is unknown how Vision Transformers scale. To address this, we scale ViT models and data, both up and down, and characterize the relationships between error rate, data, and compute. Along the way, we refine the architecture and training of ViT, reducing memory consumption and increasing accuracy of the resulting models. As a result, we successfully train a ViT model with two billion parameters, which attains a new state-of-the-art on ImageNet of 90.45% top-1 accuracy. The model also performs well on few-shot learning, for example, reaching 84.86% top-1 accuracy on ImageNet with only 10 examples per class.