Participants of the Berlin Summit on Earth Virtualization Engines (EVEs) discussed ideas and concepts to improve our ability to cope with climate change. EVEs aim to provide interactive and accessible climate simulations and data for a wide range of users. They combine high-resolution physics-based models with machine learning techniques to improve the fidelity, efficiency, and interpretability of climate projections. At their core, EVEs offer a federated data layer that enables simple and fast access to exabyte-sized climate data through simple interfaces. In this article, we summarize the technical challenges and opportunities for developing EVEs, and argue that they are essential for addressing the consequences of climate change. We are all witnessing the effects of climate change. Hotter summers, prolonged droughts, massive flooding, or ocean heat waves are examples of extreme weather and climate events that are growing in frequency and intensity. Many agree that addressing climate mitigation and adaptation is the biggest problem humanity faces today. A large group of scientists and practitioners from different climate-related domains, including some computer scientists, got together for a week in Berlin this July to discuss the concept of "Earth Virtualization Engines" (EVEs). The summit kicked off with the question: "If climate change is the most critical problem today, why are we not using the largest computers to help solve it?".

Developing algorithms for extra-terrestrial robotic exploration has always been challenging. Along with the complexity associated with these environments, one of the main issues remains the evaluation of said algorithms. With the regained interest in lunar exploration, there is also a demand for quality simulators that will enable the development of lunar robots. % In this paper, we explain how we built a Lunar simulator based on Isaac Sim, Nvidia's robotic simulator. In this paper, we propose Omniverse Lunar Robotic-Sim (OmniLRS) that is a photorealistic Lunar simulator based on Nvidia's robotic simulator. This simulation provides fast procedural environment generation, multi-robot capabilities, along with synthetic data pipeline for machine-learning applications. It comes with ROS1 and ROS2 bindings to control not only the robots, but also the environments. This work also performs sim-to-real rock instance segmentation to show the effectiveness of our simulator for image-based perception. Trained on our synthetic data, a yolov8 model achieves performance close to a model trained on real-world data, with 5% performance gap. When finetuned with real data, the model achieves 14% higher average precision than the model trained on real-world data, demonstrating our simulator's photorealism.% to realize sim-to-real. The code is fully open-source, accessible here: https://github.com/AntoineRichard/LunarSim, and comes with demonstrations.

Understanding when two pieces of text convey the same information is a goal touching many subproblems in NLP, including textual entailment and fact-checking. This problem becomes more complex when those two pieces of text are in different languages. Here, we introduce X-PARADE (Cross-lingual Paragraph-level Analysis of Divergences and Entailments), the first cross-lingual dataset of paragraph-level information divergences. Annotators label a paragraph in a target language at the span level and evaluate it with respect to a corresponding paragraph in a source language, indicating whether a given piece of information is the same, new, or new but can be inferred. This last notion establishes a link with cross-language NLI. Aligned paragraphs are sourced from Wikipedia pages in different languages, reflecting real information divergences observed in the wild. Armed with our dataset, we investigate a diverse set of approaches for this problem, including classic token alignment from machine translation, textual entailment methods that localize their decisions, and prompting of large language models. Our results show that these methods vary in their capability to handle inferable information, but they all fall short of human performance.

In aerial combat, dogfighting poses intricate challenges that demand an understanding of both strategic maneuvers and the aerodynamics of agile fighter aircraft. In this paper, we introduce TempFuser, a novel long short-term temporal fusion transformer designed to learn tactical and agile flight maneuvers in aerial dogfights. Our approach employs two distinct LSTM-based input embeddings to encode long-term sparse and short-term dense state representations. By integrating these embeddings through a transformer encoder, our model captures the tactics and agility of fighter jets, enabling it to generate end-to-end flight commands that secure dominant positions and outmaneuver the opponent. After extensive training against various types of opponent aircraft in a high-fidelity flight simulator, our model successfully learns to perform complex fighter maneuvers, consistently outperforming several baseline models. Notably, our model exhibits human-like strategic maneuvers even when facing adversaries with superior specifications, all without relying on explicit prior knowledge. Moreover, it demonstrates robust pursuit performance in challenging supersonic and low-altitude environments. Demo videos are available at https://sites.google.com/view/tempfuser.

We introduce bipol, a new metric with explainability, for estimating social bias in text data. Harmful bias is prevalent in many online sources of data that are used for training machine learning (ML) models. In a step to address this challenge we create a novel metric that involves a two-step process: corpus-level evaluation based on model classification and sentence-level evaluation based on (sensitive) term frequency (TF). After creating new models to detect bias along multiple axes using SotA architectures, we evaluate two popular NLP datasets (COPA and SQUAD). As additional contribution, we created a large dataset (with almost 2 million labelled samples) for training models in bias detection and make it publicly available. We also make public our codes.

We study the problem of Out-of-Distribution (OOD) detection, that is, detecting whether a Machine Learning (ML) model's output can be trusted at inference time. While a number of tests for OOD detection have been proposed in prior work, a formal framework for studying this problem is lacking. We propose a definition for the notion of OOD that includes both the input distribution and the ML model, which provides insights for the construction of powerful tests for OOD detection. We also propose a multiple hypothesis testing inspired procedure to systematically combine any number of different statistics from the ML model using conformal p-values. We further provide strong guarantees on the probability of incorrectly classifying an in-distribution sample as OOD. In our experiments, we find that threshold-based tests proposed in prior work perform well in specific settings, but not uniformly well across different OOD instances. In contrast, our proposed method that combines multiple statistics performs uniformly well across different datasets and neural networks architectures. Given the ubiquitous use of ML models in safety-critical applications such as self-driving and medicine, there is a need to develop methods to detect whether an ML model's output at inference time can be trusted.

If there's something that Google wants you to know, it's that the defendant in the United States' most significant antitrust trial in 25 years is not a search monopoly established through unfair, anti-competitive means--and if people get that impression, it's only because all the other search engines suck. Google is, literally, just built different. Over the first week of U.S. et al. v. Google--a suit initially filed by the U.S. Department of Justice along with more than a dozen state attorneys general in 2020--Google's lawyers put out myriad opening arguments to convince the U.S. District Court for the District of Columbia that the iconic company is not what the government accuses it of being: a search giant that reached and then stayed atop its pedestal by unfairly colluding with other tech companies to ensure its dominance. The DOJ alleges that Google intentionally crowded out search-engine competitors in order to control the sector, allowing it to overcharge advertisers, stifle the reach of other search sites, and leave consumers with no choice but to use Google's steadily degrading product. Judge Amit P. Mehta will decide, over the course of the next 10 weeks, whether the government's argument passes muster--or if Google is right that it should not be held liable under antitrust law.

On August 28th, the Deputy Secretary of Defense, Kathleen Hicks, announced what she called the Replicator initiative--an all-hands-on-deck effort to modernize the American arsenal by adding fleets of artificially intelligent, unmanned, relatively cheap weapons and equipment. She described these machines as "attritable," meaning that they can suffer attrition without compromising a mission. Imagine a swarm of hundreds or even thousands of unmanned aerial drones, communicating with each other as they collect intelligence on enemy-troop movements, and you will begin to understand the Deputy Secretary's vision for Replicator. Even if a sizable number of the drones were shot down, the information they'd gathered would have already been recorded and sent back to human operators on the ground. In one sense, Hicks's announcement, during an address titled "The Urgency to Innovate" at a meeting of National Defense Industrial Association, did not signal a wholly new approach.

Kara Frederick, tech director at the Heritage Foundation, discusses the need for regulations on artificial intelligence as lawmakers and tech titans discuss the potential risks. FIRST ON FOX: A group of 30 House Republicans is demanding to know what the Department of Justice (DOJ) is doing to combat the emergence of AI-generated child pornography on the internet. "We write to you with grave concern regarding increasing reports of artificial intelligence (AI) being used to generate child sexual abuse materials (CSAM) which are shared across the internet," Rep. Bob Good, R-Va., wrote in a letter to Attorney General Merrick Garland. "While recognizing the benefits of appropriate uses of AI, including medical research, cybersecurity defense, streamlining public transit, and may other applications, we believe action must be taken to prevent individuals from using AI to generate CSAM." Rep. Bob Good, R-Va., leads a letter to the DOJ asking about what it is doing to combat AI-generated sexually exploitative images of children.

The inaugural boss of NASA's newly-created UFO research division has been named as a former meteorologist and liaison to the Pentagon. Mark McInerney will become the US space agency's director of research into unidentified anomalous phenomena (UAPs), more commonly known as unidentified flying objects. NASA officials initially refused to reveal McInerney's identity amid fears he would be harassed, before later backtracking on the decision. Speaking about the new UFO boss following the release of the agency's highly-anticipated study into more than 800 UAP sightings, NASA's associate administrator Nicola Fox told reporters: 'They have been working there a while now, during the study, to help be a point of contact.' But when directly asked whether she could name the official, Fox replied: 'We will not give his name out.' Alien hunter: The inaugural boss of NASA's newly-created UFO research division has been named as former meteorologist and liaison to the Pentagon Mark McInerney (pictured) NASA officials initially refused to reveal McInerney's identity amid fears he would be harassed All changed hours later, however, when NASA sent out a press release that included McInerney's name and revealed he previously worked as the agency's liaison to the Pentagon.