The capabilities of large language models (LLMs) significantly depend on training data drawn from various domains. Optimizing domain-specific mixture ratios can be modeled as a bi-level optimization problem, which we simplify into a singlelevel penalized form and solve with twin networks: a proxy model trained on primary data and a dynamically updated reference model trained with additional data. Our proposed method, Twin Networks for bi-level DatA mixturE optiMization (TANDEM), measures the data efficacy through the difference between the twin models and up-weights domains that benefit more from the additional data. TANDEM provides theoretical guarantees and wider applicability, compared to prior approaches. Furthermore, our bi-level perspective suggests new settings to study domain reweighting such as data-restricted scenarios and supervised finetuning, where optimized mixture ratios significantly improve the performance.

Deep neural networks often underperform on tabular data due to sensitivity to irrelevant features and a spectral bias toward smooth, low-frequency functions, limiting their ability to capture sharp, high-frequency signals in low-label regimes. While self-supervised learning (SSL) holds promise in such settings, it remains challenging in tabular domains due to the limited availability of effective data augmentations. We introduce TANDEM (Tree-And-Neural Dual Encoder Model), a hybrid autoencoder that trains a neural encoder alongside an oblivious soft decision tree (OSDT) encoder, both guided by dedicated stochastic gating networks for sample-specific feature selection. The encoders share a decoder and are coupled via alignment losses, encouraging complementary yet consistent representations. The training-only use of the tree operates as model-based augmentation, nudging representations toward tabular-relevant features while preserving a lean inference path (only the neural encoder is deployed). Spectral analysis highlights distinct yet complementary inductive biases across encoders, and experiments on classification and regression benchmarks in low-label settings show consistent gains over strong deep, tree-based, and SSL baselines.

The capabilities of large language models (LLMs) significantly depend on training data drawn from various domains. Optimizing domain-specific mixture ratios can be modeled as a bi-level optimization problem, which we simplify into a single-level penalized form and solve with twin networks: a proxy model trained on primary data and a dynamically updated reference model trained with additional data. Our proposed method, Twin Networks for bi-level DatA mixturE optiMization (TANDEM), measures the data efficacy through the difference between the twin models and up-weights domains that benefit more from the additional data. TANDEM provides theoretical guarantees and wider applicability, compared to prior approaches. Furthermore, our bi-level perspective suggests new settings to study domain reweighting such as data-restricted scenarios and supervised fine-tuning, where optimized mixture ratios significantly improve the performance.

Handling missing data in time series classification remains a significant challenge in various domains. Traditional methods often rely on imputation, which may introduce bias or fail to capture the underlying temporal dynamics. In this paper, we propose TANDEM (Temporal Attention-guided Neural Differential Equations for Missingness), an attention-guided neural differential equation framework that effectively classifies time series data with missing values. Our approach integrates raw observation, interpolated control path, and continuous latent dynamics through a novel attention mechanism, allowing the model to focus on the most informative aspects of the data. We evaluate TANDEM on 30 benchmark datasets and a real-world medical dataset, demonstrating its superiority over existing state-of-the-art methods. Our framework not only improves classification accuracy but also provides insights into the handling of missing data, making it a valuable tool in practice.

We propose and show the efficacy of a new method to address generic inverse problems. Inverse modeling is the task whereby one seeks to determine the control parameters of a natural system that produce a given set of observed measurements. Recent work has shown impressive results using deep learning, but we note that there is a trade-off between model performance and computational time. For some applications, the computational time at inference for the best performing inverse modeling method may be overly prohibitive to its use. We present a new method that leverages multiple manifolds as a mixture of backward (e.g., inverse) models in a forward-backward model architecture. These multiple backwards models all share a common forward model, and their training is mitigated by generating training examples from the forward model. The proposed method thus has two innovations: 1) the multiple Manifold Mixture Network (MMN) architecture, and 2) the training procedure involving augmenting backward model training data using the forward model. We demonstrate the advantages of our method by comparing to several baselines on four benchmark inverse problems, and we furthermore provide analysis to motivate its design.

With much of the world now focused on the goal of net zero carbon emissions by the year 2050, industry leaders are asking: How can businesses be more sustainable? At the same time, another revolution has sprung: digital transformation. Companies are busy adopting tools and technologies to make processes more efficient and competitive. In the pursuit of global optimization, businesses have a lot to gain from thinking about these two movements in conjunction. Let's say a company with sustainability baked into its business model has decided to invest time and focus on driving more efficient operations and reducing waste.

Inspired by the human ability to perform complex manipulation in the complete absence of vision (like retrieving an object from a pocket), the robotic manipulation field is motivated to develop new methods for tactile-based object interaction. However, tactile sensing presents the challenge of being an active sensing modality: a touch sensor provides sparse, local data, and must be used in conjunction with effective exploration strategies in order to collect information. In this work, we focus on the process of guiding tactile exploration, and its interplay with task-related decision making. We propose TANDEM (TActile exploration aNd DEcision Making), an architecture to learn efficient exploration strategies in conjunction with decision making. Our approach is based on separate but co-trained modules for exploration and discrimination. We demonstrate this method on a tactile object recognition task, where a robot equipped with a touch sensor must explore and identify an object from a known set based on binary contact signals alone. TANDEM achieves higher accuracy with fewer actions than alternative methods and is also shown to be more robust to sensor noise.

Senseonics has taken their Eversense implantable CGM before an FDA panel, and Bigfoot has announced they have closed their series B $55 million financing round. The FDA has approved Dexcom's G6 system, pushing shares in the company higher in early trading. And while they didn't announce anything recently, our friends at Tandem are feeling the love as shares continue to surge higher in anticipation of their new system, which has low glucose suspend. Simply put, when it comes to diabetes devices, CGM and insulin pumps, the joint is jumping. But before everyone breaks into an unstoppable happy dance, I'd advise they slow their roll.

Digital challenger bank Tandem has turned to Personetics' artificial intelligence (AI) for personalised insights and advice. Powered by Personetics' Cognitive Banking Brain, the bank will integrate the new offerings within its digital banking. Tandem's chief executive, Ricky Knox says: "Together with our use of open banking we can use AI to get to know our customers better, predict their needs, and help them make better decisions about their money with as little effort as possible on their part. We want to do the heavy lifting for them, so they can go about their lives and not worry about day-to-day finances." Examples of personalised insights delivered by the Personetics solution include unusual spending activity; tips on how to avoid fees; and opportunities for savings or investment. Knox adds: "While AI is promising, it was important for us to work with a partner such as Personetics that has proven experience in financial services.

Artificial intelligence (AI) and the Internet of Things (IoT) have got a great deal of press in the last few years, and with good reason -- they'll both transform the way enterprises do business. And now a new Gartner report says that the synergy between them will be so powerful that enterprises should consider rolling them out in tandem. The report, "AI on the Edge: Fusing Artificial Intelligence and IoT Will Catalyze New Digital Value Creation," puts it this way: "Artificial Intelligence and the Internet of Things are symbiotic technologies that will be the foundation of a new platform for digital business value creation. CIOs engaged in IoT initiatives should leverage these capabilities for strategic advantage." The report starts out by noting that using AI and IoT in tandem will become commonplace.