Given a classification model and a prediction for some input, there are heuristic strategies for ranking features according to their importance in regard to the prediction. One common approach to this task is rooted in propositional logic and the notion of \textit{sufficient reason}. Through this concept, the categories of relevant and necessary features were proposed in order to identify the crucial aspects of the input. This paper improves the existing techniques and algorithms for deciding which are the relevant and/or necessary features, showing in particular that necessity can be detected efficiently in complex models such as neural networks. We also generalize the notion of relevancy and study associated problems. Moreover, we present a new global notion (i.e. that intends to explain whether a feature is important for the behavior of the model in general, not depending on a particular input) of \textit{usefulness} and prove that it is related to relevancy and necessity. Furthermore, we develop efficient algorithms for detecting it in decision trees and other more complex models, and experiment on three datasets to analyze its practical utility.

Realistic temporal dynamics are crucial for many video generation, processing and modelling applications, e.g. in computational fluid dynamics, weather prediction, or long-term climate simulations. Video diffusion models (VDMs) are the current state-of-the-art method for generating highly realistic dynamics. However, training VDMs from scratch can be challenging and requires large computational resources, limiting their wider application. Here, we propose a time-consistency discriminator that enables pretrained image diffusion models to generate realistic spatiotemporal dynamics. The discriminator guides the sampling inference process and does not require extensions or finetuning of the image diffusion model. We compare our approach against a VDM trained from scratch on an idealized turbulence simulation and a real-world global precipitation dataset. Our approach performs equally well in terms of temporal consistency, shows improved uncertainty calibration and lower biases compared to the VDM, and achieves stable centennial-scale climate simulations at daily time steps.

Vision language models (VLMs) are AI systems paired with both language and vision encoders to process multimodal input. They are capable of performing complex semantic tasks such as automatic captioning, but it remains an open question about how well they comprehend the visuospatial properties of scenes depicted in the images they process. We argue that descriptions of virtual objects -- objects that are not visually represented in an image -- can help test scene comprehension in these AI systems. For example, an image that depicts a person standing under a tree can be paired with the following prompt: imagine that a kite is stuck in the tree. VLMs that comprehend the scene should update their representations and reason sensibly about the spatial relations between all three objects. We describe systematic evaluations of state-of-the-art VLMs and show that their ability to process virtual objects is inadequate.

Personal use of this material is permitted. Abstract --As Large Language Models (LLMs) are widely used, understanding them systematically is key to improving their safety and realizing their full potential. Although many models are aligned using techniques such as reinforcement learning from human feedback (RLHF), they are still vulnerable to jailbreaking attacks. Some of the existing adversarial attack methods search for discrete tokens that may jailbreak a target model while others try to optimize the continuous space represented by the tokens of the model's vocabulary. While techniques based on the discrete space may prove to be inefficient, optimization of continuous token embeddings requires projections to produce discrete tokens, which might render them ineffective. T o fully utilize the constraints and the structures of the space, we develop an intrinsic optimization technique using exponentiated gradient descent with the Bregman projection method to ensure that the optimized one-hot encoding always stays within the probability simplex. We prove the convergence of the technique and implement an efficient algorithm that is effective in jailbreaking several widely used LLMs. We demonstrate the efficacy of the proposed technique using five open-source LLMs on four openly available datasets. The results show that the technique achieves a higher success rate with great efficiency compared to three other state-of-the-art jailbreaking techniques.

President Donald Trump has hailed deepening ties between the United States and the United Arab Emirates and said that the latter will invest 1.4 trillion in the former's artificial intelligence sector over the next decade. "I have absolutely no doubt that the relationship will only get bigger and better," Trump said on Thursday at a meeting with UAE President Sheikh Mohamed bin Zayed Al Nahyan, on the final leg of his three-country tour of the Gulf region that saw him strike a series of lucrative tech, business and military deals that he said amounted to 10 trillion. Sheikh Mohammed said the UAE remained "committed to working with the United States to advance peace and stability in our region and globally". The deal with UAE is expected to enable the Gulf country to build data centres vital to developing artificial intelligence models. The countries did not say which AI chips could be included in UAE data centres.

Over the course of a three-day trip to the Middle East, President Trump and his emissaries from Silicon Valley have transformed the Persian Gulf from an artificial-intelligence neophyte into an A.I. power broker. They have reached an enormous deal with the United Arab Emirates to deliver hundreds of thousands of today's most advanced chips from Nvidia annually to build one of the world's largest data center hubs in the region, three people familiar with the talks said. The shipments would begin this year, and include roughly 100,000 chips for G42, an Emirati A.I. firm, with the rest going to U.S. cloud service providers. The administration revealed the agreement on Thursday in an announcement unveiling a new A.I. campus in Abu Dhabi supported by 5 gigawatts of electrical power. It would the largest such project outside of the United States and help U.S. companies serve customers in Africa, Europe and Asia, the administration said.

Donald Trump's jaunt to the Middle East featured an entourage of billionaire tech bros, a fighter-jet escort, and business deals designed to reshape the global landscape of artificial intelligence. On the final stop of the tour in Abu Dhabi, the US President announced that unnamed US companies would partner with the United Arab Emirates to create the largest AI datacenter cluster outside of America. Trump said that the US companies will help G42, an Emirati company, build five gigawatts of AI computing capacity in the UAE. Sheikh Tahnoon bin Zayed Al Nahyan, who leads the UAE's Artificial Intelligence and Advanced Technology Council, and is in charge of a 1.5 trillion fortune aimed at building AI capabilities, said the move will strengthen the UAE's position "as a hub for cutting-edge research and sustainable development, delivering transformative benefits for humanity." A few days earlier, as Trump arrived in Riyadh, Saudi Arabia announced Humain, an AI investment firm owned by the kingdom's Public Investment Fund.

Yesterday, a user on X saw a viral post of Timothée Chalamet celebrating courtside at a Knicks game and had a simple question: Who was sitting next to him? The user tapped in Grok, X's proprietary chatbot, as people often do when they want help answering questions on the platform--the software functions like ChatGPT, except it can be summoned via reply to a post. And for the most part, Grok has performed reasonably well at providing responses. Chalamet was sitting with Kylie and Kendall Jenner, but here is how the chatbot replied: "I believe you're referring to a photo with Timothée Chalamet, but the context you mention doesn't seem to align with this image. The post discusses South African politics, which doesn't relate to Timothée or the people around him."

Breakthroughs, discoveries, and DIY tips sent every weekday. You can now own a 1 gold coin celebrating one of America's most revolutionary achievements: the NASA Space Shuttle program. The latest variant in the ongoing American Innovation 1 Coin series is available to order through the United States Mint. Selected to represent the state of Florida, the noncirculating legal tender is the third coin released this year and the 28th coin in the 15-year project first announced in 2018. While the coin's front displays the series' Statue of Liberty image, the back shows the shuttle launching above plumes of exhaust.

FIRST ON FOX: Dozens of drones that traipsed over Langley Air Force base in late 2023 revealed an astonishing oversight: Military officials did not believe they had the authority to shoot down the unmanned vehicles over the U.S. homeland. A new bipartisan bill, known as the COUNTER Act, seeks to rectify that, offering more bases the opportunity to become a "covered facility," or one that has the authority to shoot down drones that encroach on their airspace. The new bill has broad bipartisan and bicameral support, giving it a greater chance of becoming law. It's led by Armed Services Committee members Tom Cotton, R-Ark., and Kirsten Gillibrand, D-N.Y., in the Senate, and companion legislation is being introduced by August Pfluger, R-Texas, and Chrissy Houlahan, D-Pa., in the House. Currently, only half of the 360 domestic U.S. bases are considered "covered facilities" that are allowed to engage with unidentified drones.