Xenophobia is one of the key drivers of marginalisation, discrimination, and conflict, yet many prominent machine learning (ML) fairness frameworks fail to comprehensively measure or mitigate the resulting xenophobic harms. Here we aim to bridge this conceptual gap and help facilitate safe and ethical design of artificial intelligence (AI) solutions. We ground our analysis of the impact of xenophobia by first identifying distinct types of xenophobic harms, and then applying this framework across a number of prominent AI application domains, reviewing the potential interplay between AI and xenophobia on social media and recommendation systems, healthcare, immigration, employment, as well as biases in large pre-trained models. These help inform our recommendations towards an inclusive, xenophilic design of future AI systems.

This paper considers the problem of understanding the exit time for trajectories of gradient-related first-order methods from saddle neighborhoods under some initial boundary conditions. Given the 'flat' geometry around saddle points, first-order methods can struggle to escape these regions in a fast manner due to the small magnitudes of gradients encountered. In particular, while it is known that gradient-related first-order methods escape strict-saddle neighborhoods, existing analytic techniques do not explicitly leverage the local geometry around saddle points in order to control behavior of gradient trajectories. It is in this context that this paper puts forth a rigorous geometric analysis of the gradient-descent method around strict-saddle neighborhoods using matrix perturbation theory. In doing so, it provides a key result that can be used to generate an approximate gradient trajectory for any given initial conditions. In addition, the analysis leads to a linear exit-time solution for gradient-descent method under certain necessary initial conditions, which explicitly bring out the dependence on problem dimension, conditioning of the saddle neighborhood, and more, for a class of strict-saddle functions.

Discovering causal relationships from observational data is a fundamental yet challenging task. Invariant causal prediction (ICP, Peters et al., 2016) is a method for causal feature selection which requires data from heterogeneous settings and exploits that causal models are invariant. ICP has been extended to general additive noise models and to nonparametric settings using conditional independence tests. However, the latter often suffer from low power (or poor type I error control) and additive noise models are not suitable for applications in which the response is not measured on a continuous scale, but reflects categories or counts. Here, we develop transformation-model (TRAM) based ICP, allowing for continuous, categorical, count-type, and uninformatively censored responses (these model classes, generally, do not allow for identifiability when there is no exogenous heterogeneity). As an invariance test, we propose TRAM-GCM based on the expected conditional covariance between environments and score residuals with uniform asymptotic level guarantees. For the special case of linear shift TRAMs, we also consider TRAM-Wald, which tests invariance based on the Wald statistic. We provide an open-source R package 'tramicp' and evaluate our approach on simulated data and in a case study investigating causal features of survival in critically ill patients.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. A Star Wars-obsessed man who was encouraged by a chatbot "girlfriend" to slay Queen Elizabeth II was sentenced Thursday to nine years in prison for taking his plot to Windsor Castle, where he scaled the walls and was caught with a loaded crossbow on Christmas Day 2021. "I'm here to kill the queen," Jaswant Singh Chail, wearing a metal mask inspired by the dark force in the Star Wars movies, declared when he was encountered by a guard on the grounds of the castle in the early morning, according to the court. He then dropped the weapon and surrendered, and repeated his intent.

Rep. Bob Good, R-Va., joins'FOX & Friends' to discuss why he voted to oust Kevin McCarthy as speaker and how Republicans should move forward. Several House Republicans in a key bipartisan group have said they could soon see a mass exodus over their Democratic counterparts' role in Speaker Kevin McCarthy's ouster this week. "I'm really thinking strongly about leaving the Problem Solvers Caucus," Rep. Nicole Malliotakis, R-N.Y., told Fox News Digital. "I think there's a lot of Republicans who are disenchanted with the Democratic members of the Problem Solvers Caucus." McCarthy, R-Calif., became the first speaker of the House in U.S. history to be booted from the job after eight hardliners within his party joined with every Democrat to vote him out of it.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. At least seven people have reportedly been killed in Syria Thursday after a drone attack targeted a military academy graduation ceremony in Homs. "Seven dead and more than 20 regime officers wounded in a violent explosion in the area of the military academy in Homs, caused by an attack by drones during a graduation ceremony," the Syrian Observatory for Human Rights said, according to the AFP. The source of the attack is unclear.

A drone attack on a military college in Syria's Homs province during a graduation ceremony has killed and wounded civilians and military personnel, according to Syrian officials. Drones laden with explosives targeted the ceremony on Thursday as it came to an end, Syria's military said in a statement. They accused fighters "backed by known international forces" for the attack. The statement did not specify the number of casualties but said some of the wounded were in critical condition, including women and children. There was no immediate claim of responsibility.

It's finally nearing the end of a month filled with consumer tech announcements, and Wednesday's Google event felt like the grand finale. While Google only sells a fraction of the number of phones and smartwatches pumped out by Apple and Samsung, the company's work in mobile software, large language models, productivity services, and computational photography make it just as much of a heavyweight when it comes to consumer tech. But Google's reach also extends far beyond your pocket and your wrist. Let us not forget about the company's dominance in search. In fact, it's currently in the throes of a protracted antitrust trial brought by the US government.

An organization dedicated to the safe development of artificial intelligence released a "breakthrough paper" it said will help humans better control the technology as it spreads. "We can't trust AIs if we don't know what they are thinking or how they work on the inside," Dan Hendrycks, director of the Center for AI Safety, told Fox News Digital. Hendrycks made the comments after the Center for AI Safety (CAIS) released a paper this week diving into the inner workings of the mind of AI systems, looking for ways that humans could better understand and control and understand AI technologies and mitigate some of the risks they pose. META MAY BE USING YOUR FACEBOOK, INSTAGRAM TO'FEED THE BEAST' OF NEW TECH According to the CAIS, the paper demonstrated ways humans can control and detect when AI systems are telling truths or lies, when they behave morally or immorally, whether they act with emotions such as anger, fear and joy, and how to make them less biased. The paper also looked at ways to develop systems that can resist jailbreaks, a practice where users can exploit vulnerabilities in AI systems and potentially use them outside desired protocols.

Artificial Neural Networks (ANNs) trained with Backpropagation (BP) show astounding performance and are increasingly often used in performing our daily life tasks. However, ANNs are highly vulnerable to adversarial attacks, which alter inputs with small targeted perturbations that drastically disrupt the models' performance. The most effective method to make ANNs robust against these attacks is adversarial training, in which the training dataset is augmented with exemplary adversarial samples. Unfortunately, this approach has the drawback of increased training complexity since generating adversarial samples is very computationally demanding. In contrast to ANNs, humans are not susceptible to adversarial attacks. Therefore, in this work, we investigate whether biologically-plausible learning algorithms are more robust against adversarial attacks than BP. In particular, we present an extensive comparative analysis of the adversarial robustness of BP and Present the Error to Perturb the Input To modulate Activity (PEPITA), a recently proposed biologically-plausible learning algorithm, on various computer vision tasks. We observe that PEPITA has higher intrinsic adversarial robustness and, with adversarial training, has a more favourable natural-vs-adversarial performance trade-off as, for the same natural accuracies, PEPITA's adversarial accuracies decrease in average by 0.26% and BP's by 8.05%.