Indoor localization systems have become increasingly important in a wide range of applications, including industry, security, logistics, and emergency services. However, the growing demand for accurate localization has heightened concerns over privacy, as many localization systems rely on active signals that can be misused by an adversary to track users' movements or manipulate their measurements. This paper presents PassiFi, a novel passive Wi-Fi time-based indoor localization system that effectively balances accuracy and privacy. PassiFi uses a passive WiFi Time Difference of Arrival (TDoA) approach that ensures users' privacy and safeguards the integrity of their measurement data while still achieving high accuracy. The system adopts a fingerprinting approach to address multi-path and non-line-of-sight problems and utilizes deep neural networks to learn the complex relationship between TDoA and location. Evaluation in a real-world testbed demonstrates PassiFi's exceptional performance, surpassing traditional multilateration by 128%, achieving sub-meter accuracy on par with state-of-the-art active measurement systems, all while preserving privacy.

Neural network models are one of the most successful approaches to machine learning, enjoying an enormous amount of development and research over recent years and finding concrete real-world applications in almost any conceivable area of science, engineering and modern life in general. The theoretical understanding of neural networks trails significantly behind their practical success and the engineering heuristics that have grown up around them. Random matrix theory provides a rich framework of tools with which aspects of neural network phenomenology can be explored theoretically. In this thesis, we establish significant extensions of prior work using random matrix theory to understand and describe the loss surfaces of large neural networks, particularly generalising to different architectures. Informed by the historical applications of random matrix theory in physics and elsewhere, we establish the presence of local random matrix universality in real neural networks and then utilise this as a modeling assumption to derive powerful and novel results about the Hessians of neural network loss surfaces and their spectra. In addition to these major contributions, we make use of random matrix models for neural network loss surfaces to shed light on modern neural network training approaches and even to derive a novel and effective variant of a popular optimisation algorithm. Overall, this thesis provides important contributions to cement the place of random matrix theory in the theoretical study of modern neural networks, reveals some of the limits of existing approaches and begins the study of an entirely new role for random matrix theory in the theory of deep learning with important experimental discoveries and novel theoretical results based on local random matrix universality.

This paper describes our submission to the MEDIQA-Chat 2023 shared task for automatic clinical note generation from doctor-patient conversations. We report results for two approaches: the first fine-tunes a pre-trained language model (PLM) on the shared task data, and the second uses few-shot in-context learning (ICL) with a large language model (LLM). Both achieve high performance as measured by automatic metrics (e.g. ROUGE, BERTScore) and ranked second and first, respectively, of all submissions to the shared task. Expert human scrutiny indicates that notes generated via the ICL-based approach with GPT-4 are preferred about as often as human-written notes, making it a promising path toward automated note generation from doctor-patient conversations.

We propose a dynamic graph representation method, showcasing its rich representational capacity and establishing some of its theoretical properties. Our representation falls under the bind-and-sum approach in hyperdimensional computing (HDC), and we show that the tensor product is the most general binding operation that respects the superposition principle employed in HDC. We also establish some precise results characterizing the behavior of our method, including a memory vs. size analysis of how our representation's size must scale with the number of edges in order to retain accurate graph operations. True to its HDC roots, we also compare our graph representation to another typical HDC representation, the Hadamard-Rademacher scheme, showing that these two graph representations have the same memory-capacity scaling. We establish a link to adjacency matrices, showing that our method is a pseudo-orthogonal generalization of adjacency matrices. In light of this, we briefly discuss its applications toward a dynamic compressed representation of large sparse graphs.

A well-known limitation of existing molecular generative models is that the generated molecules highly resemble those in the training set. To generate truly novel molecules that may have even better properties for de novo drug discovery, more powerful exploration in the chemical space is necessary. To this end, we propose Molecular Out-Of-distribution Diffusion(MOOD), a score-based diffusion scheme that incorporates out-of-distribution (OOD) control in the generative stochastic differential equation (SDE) with simple control of a hyperparameter, thus requires no additional costs. Since some novel molecules may not meet the basic requirements of real-world drugs, MOOD performs conditional generation by utilizing the gradients from a property predictor that guides the reverse-time diffusion process to high-scoring regions according to target properties such as protein-ligand interactions, drug-likeness, and synthesizability. This allows MOOD to search for novel and meaningful molecules rather than generating unseen yet trivial ones. We experimentally validate that MOOD is able to explore the chemical space beyond the training distribution, generating molecules that outscore ones found with existing methods, and even the top 0.01% of the original training pool. Our code is available at https://github.com/SeulLee05/MOOD.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. Extremist rebels in eastern Congo killed at least nine people with knives and guns, a civil society organization said Friday. The attack happened Thursday evening on the Kyondo-Kyavinyonge road in North Kivu province, said Meleki Mulala the coordinator for the Congolese civil society group for the Ruwenzori sector. Civilians were taken from their homes before they were killed, and many homes were looted, he said.

Fox News' Eben Brown reports on how more companies are using A.I. technology to set retail prices based on data-driven supply-and-demand. A new artificial intelligence (AI) tool could help African countries better track and predict crop rotations and yields, providing a key tool to help mitigate food crisis across the continent. "Relying on conventional analytic techniques alone will not deliver the effective decision-making we need to meet these challenges," Racine Ly, the director of data management for the project, told Science X. "Since this is data that researchers and decision-makers most importantly will use to make decisions, we needed to make sure that the data is correct, and the predictions are accurate," he added. AKADEMIYA2063, a research organization, said the Africa Agriculture Watch (AAgWa) tool will help prioritize and maximize the production of staple foods such as maize, cassava and sorghum. WILL AI EVER BE SMART ENOUGH TO DECIPHER FEDERAL REGULATIONS?

Retrieval-augmented language models such as Fusion-in-Decoder are powerful, setting the state of the art on a variety of knowledge-intensive tasks. However, they are also expensive, due to the need to encode a large number of retrieved passages. Some work avoids this cost by pre-encoding a text corpus into a memory and retrieving dense representations directly. However, pre-encoding memory incurs a severe quality penalty as the memory representations are not conditioned on the current input. We propose LUMEN, a hybrid between these two extremes, pre-computing the majority of the retrieval representation and completing the encoding on the fly using a live encoder that is conditioned on the question and fine-tuned for the task. We show that LUMEN significantly outperforms pure memory on multiple question-answering tasks while being much cheaper than FiD, and outperforms both for any given compute budget. Moreover, the advantage of LUMEN over FiD increases with model size.

Visual Question Answering (VQA) is a multi-disciplinary artificial intelligence research problem that has attracted the attention of researchers from computer vision, natural language processing, knowledge representation, and other machine learning communities. To solve that question, VQA is a task of generating natural language answers when a question in natural language is asked related to an image. In recent years, visual question answering as a result of the flourish in this field, datasets, metrics, and models have been proposed, and the scope of research has been expanded. Although artificial intelligence has solved several different problems, such as image classification and natural language processing (NLP), it is hard to model a problem which needs different types of data. For instance, mixing computer vision with NLP to retrieve some information about an image from a question has tricked researchers for several years.

Earth System Models (ESMs) are the primary tools for investigating future Earth system states at time scales from decades to centuries, especially in response to anthropogenic greenhouse gas release. State-of-the-art ESMs can reproduce the observational global mean temperature anomalies of the last 150 years. Nevertheless, ESMs need further improvements, most importantly regarding (i) the large spread in their estimates of climate sensitivity, i.e., the temperature response to increases in atmospheric greenhouse gases, (ii) the modeled spatial patterns of key variables such as temperature and precipitation, (iii) their representation of extreme weather events, and (iv) their representation of multistable Earth system components and their ability to predict associated abrupt transitions. Here, we argue that making ESMs automatically differentiable has huge potential to advance ESMs, especially with respect to these key shortcomings. First, automatic differentiability would allow objective calibration of ESMs, i.e., the selection of optimal values with respect to a cost function for a large number of free parameters, which are currently tuned mostly manually. Second, recent advances in Machine Learning (ML) and in the amount, accuracy, and resolution of observational data promise to be helpful with at least some of the above aspects because ML may be used to incorporate additional information from observations into ESMs. Automatic differentiability is an essential ingredient in the construction of such hybrid models, combining process-based ESMs with ML components. We document recent work showcasing the potential of automatic differentiation for a new generation of substantially improved, data-informed ESMs.