Recent advances in recommendation scaling laws have led to foundation models of unprecedented complexity. While these models offer superior performance, their computational demands make real-time serving impractical, often forcing practitioners to rely on knowledge distillation—compromising serving quality for efficiency. To address this challenge, we present SOLARIS (Speculative Offloading of Latent-bAsed Representation for Inference Scaling), a novel framework inspired by speculative decoding. SOLARIS proactively precomputes user-item interaction embeddings by predicting which user-item pairs are likely to appear in future requests, and asynchronously generating their foundation model representations ahead of time. This approach decouples the costly foundation model inference from the latency-critical serving path, enabling real-time knowledge transfer from models previously considered too expensive for online use. Deployed across Meta's advertising system serving billions of daily requests, SOLARIS achieves 0.67% revenue-driving top-line metrics gain, demonstrating its effectiveness at scale.

Existing deep learning-based surrogate models facilitate efficient data generation, but fall short in uncertainty quantification, efficient parameter space exploration, and reverse prediction. In our work, we introduce SurroFlow, a novel normalizing flow-based surrogate model, to learn the invertible transformation between simulation parameters and simulation outputs. The model not only allows accurate predictions of simulation outcomes for a given simulation parameter but also supports uncertainty quantification in the data generation process. Additionally, it enables efficient simulation parameter recommendation and exploration. We integrate SurroFlow and a genetic algorithm as the backend of a visual interface to support effective user-guided ensemble simulation exploration and visualization. Our framework significantly reduces the computational costs while enhancing the reliability and exploration capabilities of scientific surrogate models.

Although many deep-learning-based super-resolution approaches have been proposed in recent years, because no ground truth is available in the inference stage, few can quantify the errors and uncertainties of the super-resolved results. For scientific visualization applications, however, conveying uncertainties of the results to scientists is crucial to avoid generating misleading or incorrect information. In this paper, we propose PSRFlow, a novel normalizing flow-based generative model for scientific data super-resolution that incorporates uncertainty quantification into the super-resolution process. PSRFlow learns the conditional distribution of the high-resolution data based on the low-resolution counterpart. By sampling from a Gaussian latent space that captures the missing information in the high-resolution data, one can generate different plausible super-resolution outputs. The efficient sampling in the Gaussian latent space allows our model to perform uncertainty quantification for the super-resolved results. During model training, we augment the training data with samples across various scales to make the model adaptable to data of different scales, achieving flexible super-resolution for a given input. Our results demonstrate superior performance and robust uncertainty quantification compared with existing methods such as interpolation and GAN-based super-resolution networks.

Deep learning based latent representations have been widely used for numerous scientific visualization applications such as isosurface similarity analysis, volume rendering, flow field synthesis, and data reduction, just to name a few. However, existing latent representations are mostly generated from raw data in an unsupervised manner, which makes it difficult to incorporate domain interest to control the size of the latent representations and the quality of the reconstructed data. In this paper, we present a novel importance-driven latent representation to facilitate domain-interest-guided scientific data visualization and analysis. We utilize spatial importance maps to represent various scientific interests and take them as the input to a feature transformation network to guide latent generation. We further reduced the latent size by a lossless entropy encoding algorithm trained together with the autoencoder, improving the storage and memory efficiency. We qualitatively and quantitatively evaluate the effectiveness and efficiency of latent representations generated by our method with data from multiple scientific visualization applications.

Generating compact and effective numerical representations of data is a fundamental step for many machine learning tasks. Traditionally, handcrafted features are used but as deep learning starts to show its potential, using deep learning models to extract compact representations becomes a new trend. Among them, adopting vectors from the model’s latent space is the most popular. There are several studies focused on visual analysis of latent space in NLP and computer vision . However, relatively little work has been done for music information retrieval (MIR) especially incorporating visualization. To bridge this gap, we propose a visual analysis system utilizing Autoencoders to facilitate analysis and exploration of traditional Chinese music. Due to the lack of proper traditional Chinese music data, we construct a labeled dataset from a collection of pre-recorded audios and then convert them into spectrograms . Our system takes music features learned from two deep learning models (a fully-connected Autoencoder and a Long Short-Term Memory (LSTM) Autoencoder) as input. Through interactive selection, similarity calculation, clustering and listening, we show that the latent representations of the encoded data allow our system to identify essential music elements, which lay the foundation for further analysis and retrieval of Chinese music in the future.

With more sensors embedded and functions added, mobile phones tend to be more critical to daily life. Researchers have been using the sensor data to recognize human activity these days; meanwhile, the mobile application usage prediction is also gradually brought into the spotlight. In this paper, we leveraged a state-of-the-art technique, which is LSTM, to model the mobile application usage data, also introduced a data fusion technique that eventually accomplished an over 90% of prediction accuracy. To validate the generality of our proposed solution, we applied the model on a public dataset. Our proposed solution treated the mobile application usage as a time series problem which is novel in the related field; it has the advantages of low resource consumption, short training time, as well as a generality. With the growth of users' reliance on mobile phones, mobile application usage prediction will be more useful in the future.

In the big data era, image processing for large dataset becomes an issue that requires immediate solution. We proposed an effective solution for training a deep convolutional neural network on Apache Spark, successfully reduced the processing time by nearly a half also retained a high recognition accuracy at the meantime. This network model could also prevent overfitting by applying dropout algorithm. Experiments are performed on MNIST dataset to make comparisons with different Convolutional Neural Networks (CNN) architectures in multiple dimensions thoroughly.