This paper proposes an unmanned aerial vehicle (UAV) aided content management system in communication-challenged disaster scenarios. Without cellular infrastructure in such scenarios, community of stranded users can be provided access to situation-critical contents using a hybrid network of static and traveling UAVs. A set of relatively static anchor UAVs can download content from central servers and provide content access to its local users. A set of ferrying UAVs with wider mobility can provision content to users by shuffling them across different anchor UAVs while visiting different communities of users. The objective is to design a content dissemination system that on-the-fly learns content caching policies for maximizing content availability to the stranded users. This paper proposes a decentralized Top-k Multi-Armed Bandit Learning model for UAV-caching decision-making that takes geo-temporal differences in content popularity and heterogeneity in content demands into consideration. The proposed paradigm is able to combine the expected reward maximization attribute and a proposed multi-dimensional reward structure of Top-k Multi-Armed Bandit, for caching decision at the UAVs. This study is done for different user-specified tolerable access delay, heterogeneous popularity distributions, and inter-community geographical characteristics. Functional verification and performance evaluation of the proposed caching framework is done for a wide range of network size, UAV distribution, and content popularity.

Urban environments, characterized by their complex, multi-layered networks encompassing physical, social, economic, and environmental dimensions, face significant challenges in the face of rapid urbanization. These challenges, ranging from traffic congestion and pollution to social inequality, call for advanced technological interventions. Recent developments in big data, artificial intelligence, urban computing, and digital twins have laid the groundwork for sophisticated city modeling and simulation. However, a gap persists between these technological capabilities and their practical implementation in addressing urban challenges in an systemic-intelligent way. This paper proposes Urban Generative Intelligence (UGI), a novel foundational platform integrating Large Language Models (LLMs) into urban systems to foster a new paradigm of urban intelligence. UGI leverages CityGPT, a foundation model trained on city-specific multi-source data, to create embodied agents for various urban tasks. These agents, operating within a textual urban environment emulated by city simulator and urban knowledge graph, interact through a natural language interface, offering an open platform for diverse intelligent and embodied agent development. This platform not only addresses specific urban issues but also simulates complex urban systems, providing a multidisciplinary approach to understand and manage urban complexity. This work signifies a transformative step in city science and urban intelligence, harnessing the power of LLMs to unravel and address the intricate dynamics of urban systems. The code repository with demonstrations will soon be released here https://github.com/tsinghua-fib-lab/UGI.

This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that our most-capable Gemini Ultra model advances the state of the art in 30 of 32 of these benchmarks - notably being the first model to achieve human-expert performance on the well-studied exam benchmark MMLU, and improving the state of the art in every one of the 20 multimodal benchmarks we examined. We believe that the new capabilities of Gemini models in cross-modal reasoning and language understanding will enable a wide variety of use cases and we discuss our approach toward deploying them responsibly to users.

While there has been significant progress in curriculum learning and continuous learning for training agents to generalize across a wide variety of environments in the context of single-agent reinforcement learning, it is unclear if these algorithms would still be valid in a multi-agent setting. In a competitive setting, a learning agent can be trained by making it compete with a curriculum of increasingly skilled opponents. However, a general intelligent agent should also be able to learn to act around other agents and cooperate with them to achieve common goals. When cooperating with other agents, the learning agent must (a) learn how to perform the task (or subtask), and (b) increase the overall team reward. In this paper, we aim to answer the question of what kind of cooperative teammate, and a curriculum of teammates should a learning agent be trained with to achieve these two objectives. Our results on the game Overcooked show that a pre-trained teammate who is less skilled is the best teammate for overall team reward but the worst for the learning of the agent. Moreover, somewhat surprisingly, a curriculum of teammates with decreasing skill levels performs better than other types of curricula.

MineObserver 2.0 is an AI framework that uses Computer Vision and Natural Language Processing for assessing the accuracy of learner-generated descriptions of Minecraft images that include some scientifically relevant content. The system automatically assesses the accuracy of participant observations, written in natural language, made during science learning activities that take place in Minecraft. We demonstrate our system working in real-time and describe a teacher support dashboard to showcase observations, both of which advance our previous work. We present the results of a study showing that MineObserver 2.0 improves over its predecessor both in perceived accuracy of the system's generated descriptions as well as in usefulness of the system's feedback. In future work we intend improve system-generated descriptions, give teachers more control and upgrade the system to perform continuous learning to more effectively and rapidly respond to novel observations made by learners.

In many real-world applications, from robotics to pedestrian trajectory prediction, there is a need to predict multiple real-valued outputs to represent several potential scenarios. Current deep learning techniques to address multiple-output problems are based on two main methodologies: (1) mixture density networks, which suffer from poor stability at high dimensions, or (2) multiple choice learning (MCL), an approach that uses $M$ single-output functions, each only producing a point estimate hypothesis. This paper presents a Mixture of Multiple-Output functions (MoM) approach using a novel variant of dropout, Multiple Hypothesis Dropout. Unlike traditional MCL-based approaches, each multiple-output function not only estimates the mean but also the variance for its hypothesis. This is achieved through a novel stochastic winner-take-all loss which allows each multiple-output function to estimate variance through the spread of its subnetwork predictions. Experiments on supervised learning problems illustrate that our approach outperforms existing solutions for reconstructing multimodal output distributions. Additional studies on unsupervised learning problems show that estimating the parameters of latent posterior distributions within a discrete autoencoder significantly improves codebook efficiency, sample quality, precision and recall.

In recent years, the rapid advancement and impressive capabilities of Large Language Models (LLMs) have been evident across various domains. This paper explores the application, implications, and potential of LLMs in building energy efficiency and decarbonization studies. The wide-ranging capabilities of LLMs are examined in the context of the building energy field, including intelligent control systems, code generation, data infrastructure, knowledge extraction, and education. Despite the promising potential of LLMs, challenges including complex and expensive computation, data privacy, security and copyright, complexity in fine-tuned LLMs, and self-consistency are discussed. The paper concludes with a call for future research focused on the enhancement of LLMs for domain-specific tasks, multi-modal LLMs, and collaborative research between AI and energy experts.

Diffusion models have demonstrated strong potential for robotic trajectory planning. However, generating coherent and long-horizon trajectories from high-level instructions remains challenging, especially for complex tasks requiring multiple sequential skills. We propose SkillDiffuser, an end-to-end hierarchical planning framework integrating interpretable skill learning with conditional diffusion planning to address this problem. At the higher level, the skill abstraction module learns discrete, human-understandable skill representations from visual observations and language instructions. These learned skill embeddings are then used to condition the diffusion model to generate customized latent trajectories aligned with the skills. It allows for generating diverse state trajectories that adhere to the learnable skills. By integrating skill learning with conditional trajectory generation, SkillDiffuser produces coherent behavior following abstract instructions across diverse tasks. Experiments on multi-task robotic manipulation benchmarks like Meta-World and LOReL demonstrate state-of-the-art performance and human-interpretable skill representations from SkillDiffuser.

The increasing demand for wireless communication underscores the need to optimize radio frequency spectrum utilization. An effective strategy for leveraging underutilized licensed frequency bands is cooperative spectrum sensing (CSS), which enable multiple secondary users (SUs) to collaboratively detect the spectrum usage of primary users (PUs) prior to accessing the licensed spectrum. The increasing popularity of machine learning has led to a shift from traditional CSS methods to those based on deep learning. However, deep learning-based CSS methods often rely on centralized learning, posing challenges like communication overhead and data privacy risks. Recent research suggests vertical federated learning (VFL) as a potential solution, with its core concept centered on partitioning the deep neural network into distinct segments, with each segment is trained separately. However, existing VFL-based CSS works do not fully address the practical challenges arising from streaming data and the objective shift. In this work, we introduce online vertical federated learning (OVFL), a robust framework designed to address the challenges of ongoing data stream and shifting learning goals. Our theoretical analysis reveals that OVFL achieves a sublinear regret bound, thereby evidencing its efficiency. Empirical results from our experiments show that OVFL outperforms benchmarks in CSS tasks. We also explore the impact of various parameters on the learning performance.

Self-supervised learning enables the training of large neural models without the need for large, labeled datasets. It has been generating breakthroughs in several fields, including computer vision, natural language processing, biology, and speech. In particular, the state-of-the-art in several speech processing applications, such as automatic speech recognition or speaker identification, are models where the latent representation is learned using self-supervised approaches. Several configurations exist in self-supervised learning for speech, including contrastive, predictive, and multilingual approaches. There is, however, a crucial limitation in most existing approaches: their high computational costs. These costs limit the deployment of models, the size of the training dataset, and the number of research groups that can afford research with large self-supervised models. Likewise, we should consider the environmental costs that high energy consumption implies. Efforts in this direction comprise optimization of existing models, neural architecture efficiency, improvements in finetuning for speech processing tasks, and data efficiency. But despite current efforts, more work could be done to address high computational costs in self-supervised representation learning.