Recent research has shown that the integration of Reinforcement Learning (RL) with Moving Target Defense (MTD) can enhance cybersecurity in Internet-of-Things (IoT) devices. Nevertheless, the practicality of existing work is hindered by data privacy concerns associated with centralized data processing in RL, and the unsatisfactory time needed to learn right MTD techniques that are effective against a rising number of heterogeneous zero-day attacks. Thus, this work presents CyberForce, a framework that combines Federated and Reinforcement Learning (FRL) to collaboratively and privately learn suitable MTD techniques for mitigating zero-day attacks. CyberForce integrates device fingerprinting and anomaly detection to reward or penalize MTD mechanisms chosen by an FRL-based agent. The framework has been deployed and evaluated in a scenario consisting of ten physical devices of a real IoT platform affected by heterogeneous malware samples. A pool of experiments has demonstrated that CyberForce learns the MTD technique mitigating each attack faster than existing RL-based centralized approaches. In addition, when various devices are exposed to different attacks, CyberForce benefits from knowledge transfer, leading to enhanced performance and reduced learning time in comparison to recent works. Finally, different aggregation algorithms used during the agent learning process provide CyberForce with notable robustness to malicious attacks.

Legged robots leverage ground contacts and the reaction forces they provide to achieve agile locomotion. However, uncertainty coupled with contact discontinuities can lead to failure, especially in real-world environments with unexpected height variations such as rocky hills or curbs. To enable dynamic traversal of extreme terrain, this work introduces 1) a proprioception-based gait planner for estimating unknown hybrid events due to elevation changes and responding by modifying contact schedules and planned footholds online, and 2) a two-degree-of-freedom tail for improving contact-independent control and a corresponding decoupled control scheme for better versatility and efficiency. Simulation results show that the gait planner significantly improves stability under unforeseen terrain height changes compared to methods that assume fixed contact schedules and footholds. Further, tests have shown that the tail is particularly effective at maintaining stability when encountering a terrain change with an initial angular disturbance. The results show that these approaches work synergistically to stabilize locomotion with elevation changes up to 1.5 times the leg length and tilted initial states.

We present COVID-Q, a set of 1,690 questions about COVID-19 from 13 sources, which we annotate into 15 question categories and 207 question clusters. The most common questions in our dataset asked about transmission, prevention, and societal effects of COVID, and we found that many questions that appeared in multiple sources were not answered by any FAQ websites of reputable organizations such as the CDC and FDA. We post our dataset publicly at https://github.com/JerryWeiAI/COVID-Q. For classifying questions into 15 categories, a BERT baseline scored 58.1% accuracy when trained on 20 examples per category, and for a question clustering task, a BERT + triplet loss baseline achieved 49.5% accuracy. We hope COVID-Q can help either for direct use in developing applied systems or as a domain-specific resource for model evaluation.

A fundamental question in theoretical machine learning is generalization. Over the past decades, the PAC-Bayesian approach has been established as a flexible framework to address the generalization capabilities of machine learning algorithms, and design new ones. Recently, it has garnered increased interest due to its potential applicability for a variety of learning algorithms, including deep neural networks. In parallel, an information-theoretic view of generalization has developed, wherein the relation between generalization and various information measures has been established. This framework is intimately connected to the PAC-Bayesian approach, and a number of results have been independently discovered in both strands. In this monograph, we highlight this strong connection and present a unified treatment of generalization. We present techniques and results that the two perspectives have in common, and discuss the approaches and interpretations that differ. In particular, we demonstrate how many proofs in the area share a modular structure, through which the underlying ideas can be intuited. We pay special attention to the conditional mutual information (CMI) framework; analytical studies of the information complexity of learning algorithms; and the application of the proposed methods to deep learning. This monograph is intended to provide a comprehensive introduction to information-theoretic generalization bounds and their connection to PAC-Bayes, serving as a foundation from which the most recent developments are accessible. It is aimed broadly towards researchers with an interest in generalization and theoretical machine learning.

Variational inference is an approximation framework for Bayesian inference that seeks to improve quantified uncertainty in predictions by optimizing a simplified distribution over parameters to stand in for the full posterior. Capturing model variations that remain consistent with training data enables more robust predictions by reducing parameter sensitivity. This work introduces a fixed-point optimization for variational inference that is applicable when every feasible log density can be expressed as a linear combination of functions from a given basis. In such cases, the optimizer becomes a fixed-point of projective integral updates. When the basis spans univariate quadratics in each parameter, feasible densities are Gaussian and the projective integral updates yield quasi-Newton variational Bayes (QNVB). Other bases and updates are also possible. As these updates require high-dimensional integration, this work first proposes an efficient quasirandom quadrature sequence for mean-field distributions. Each iterate of the sequence contains two evaluation points that combine to correctly integrate all univariate quadratics and, if the mean-field factors are symmetric, all univariate cubics. More importantly, averaging results over short subsequences achieves periodic exactness on a much larger space of multivariate quadratics. The corresponding variational updates require 4 loss evaluations with standard (not second-order) backpropagation to eliminate error terms from over half of all multivariate quadratic basis functions. This integration technique is motivated by first proposing stochastic blocked mean-field quadratures, which may be useful in other contexts. A PyTorch implementation of QNVB allows for better control over model uncertainty during training than competing methods. Experiments demonstrate superior generalizability for multiple learning problems and architectures.

Elon Musk ordered his Starlink satellite communications network to be turned off near the Crimean coast last year to hobble a Ukrainian drone attack on Russian warships, according to a new biography. CNN quoted an excerpt from the biography Elon Musk by Walter Isaacson, which described how armed submarine drones were approaching their targets when they "lost connectivity and washed ashore harmlessly". The biography, due out on Tuesday, alleges Musk ordered Starlink engineers to turn off service in the area of the attack because of his concern that Vladimir Putin would respond with nuclear weapons to a Ukrainian attack on Russian-occupied Crimea. He is reported to have said that Ukraine was "going too far" in threatening to inflict a "strategic defeat" on the Kremlin. Musk's threats to withdraw Starlink communications at various stages of the conflict have been previously reported, but this is the first time it has been alleged he cut off Ukrainian forces in the middle of a specific operation.

The news: One of the leading companies offering alternatives to lithium batteries for the grid has just received a nearly $400 million loan from the US Department of Energy. Eos Energy makes zinc-halide batteries, which the firm hopes could one day be used to store renewable energy at a lower cost than is possible with existing lithium-ion batteries. What they're made of: Eos's batteries are primarily made from zinc, the fourth most produced metal in the world, and use a water-based electrolyte (the liquid that moves charge around in a battery) instead of organic solvent. This makes them more stable than lithium-ion cells, and means they won't catch fire. Why it matters: While the cost of lithium-ion batteries has plummeted over the past decade, there's a growing need for even cheaper options.

What is unique about AI is also what is most feared and celebrated--its ability to match some of our own skills, and then to go further, accomplishing what humans cannot. AI's capacity to model itself on human behavior has become its defining feature. Yet behind every advance in machine learning and large language models are, in fact, people--both the often obscured human labor that makes large language models safer to use, and the individuals who make critical decisions on when and how to best use this technology. Reporting on people and influence is what TIME does best. That led us to the TIME100 AI.

Fox News correspondent Gillian Turner has the latest on the president's focus amid calls for an impeachment inquiry on'Special Report.' The top Republican on the Senate Energy Committee will warn Thursday against allowing U.S. artificial intelligence capabilities to fall into China's hands when the panel meets for a hearing on the topic. Senators returned to Capitol Hill just days ago after spending the month of August in their home states. AI is expected to be a prominent topic for lawmakers as they race to get ahead of the rapidly advancing technology. It's also the topic at the heart of Thursday's hearing led by Energy Committee Chair Joe Manchin, D-W.Va., and ranking member John Barrasso, R-Wyo., that aims to examine how AI has affected the U.S. energy sector and how the federal government can stay competitive in that lane.

Looking at this new picture from NASA's Perseverance rover, you'd be forgiven for thinking there's something fishy afoot on the Red Planet. That's because the car-sized robot has snapped an image of two separate boulders resembling a shark fin and a crab claw. The US space agency shared this latest discovery on X (formerly known as Twitter), prompting a wave of replies from space fans who joked that the crab-like rock was the remains of the'Almighty Great Cosmic Crab'. Others said the'claw' looked more like a coffee bean or the head of a turtle'digging a hole for its eggs', while some quipped that the shark fin might actually be the'back plates' of a Stegosaurus. The photos, which were taken last month, are the latest example of a phenomenon known as pareidolia -- where the human brain wants to make sense of what the eyes see so creates a meaning which isn't real. Peculiar: NASA's Perseverance rover has snapped images of two separate boulders resembling a shark fin and a crab claw Most famously with Mars, this happened in 1976 when NASA's Viking 1 spacecraft captured an image of what looked like a face carved into the surface of the Red Planet.