Valve's Steam Machine: Everything we know Representatives want a record of all related communications. Several hundred protesters had gathered near the Broadview ICE center, chanting against immigration enforcement policies. The House Judiciary Committee wants the US Department of Justice to turn over all its communications with both Apple and Google regarding the companies' decisions to remove apps that shared information about sightings of US Immigration and Customs Enforcement officers. Several apps were removed from both Apple's App Store and Google's Play Store in October. "The coercion and censorship campaign, which ultimately targets the users of ICE-monitoring applications, is a clear effort to silence this Administration's critics and suppress any evidence that would expose the Administration's lies, including its Orwellian attempts to cover up the murders of Renee and Alex," Raskin wrote to Bondi.

Ferrari's New Jony Ive-Designed EV Is Swathed in Glass and Aluminum We got a peek at the interior of Ferrari's new Luce electric car, which was dreamed up by famed ex-Apple designer and his firm, LoveFrom. It looks and feels a whole lot like an Apple product. Despite Ferrari dramatically scaling back its EV plans at the end of 2025, it's no exaggeration to say that the reveal of the Italian automaker's first full electric car is going to be automotive event of 2026. While the exterior is still under wraps, Ferrari has unveiled the interior of its upcoming electric vehicle designed by LoveFrom, the creative firm of Apple's former chief designer, Jony Ive. It may not turn out quite like the Project Titan car Apple worked on for a decade then killed in 2024, but it sure does look like it has similar DNA. "We are entering a new era in Ferrari," the company's CEO Benedetto Vigna said at the unveiling, which took place last week at San Francisco's pyramid-shaped Transamerica building.

The advent of Federated Learning (FL) highlights the practical necessity for the'right to be forgotten' for all clients, allowing them to request data deletion from the machine learning model's service provider. This necessity has spurred a growing demand for Federated Unlearning (FU). Feature unlearning has gained considerable attention due to its applications in unlearning sensitive, backdoor, and biased features. Existing methods employ the influence function to achieve feature unlearning, which is impractical for FL as it necessitates the participation of other clients, if not all, in the unlearning process. Furthermore, current research lacks an evaluation of the effectiveness of feature unlearning. To address these limitations, we define feature sensitivity in evaluating feature unlearning according to Lipschitz continuity. This metric characterizes the model output's rate of change or sensitivity to perturbations in the input feature. We then propose an effective federated feature unlearning framework called Ferrari, which minimizes feature sensitivity. Extensive experimental results and theoretical analysis demonstrate the effectiveness of Ferrari across various feature unlearning scenarios, including sensitive, backdoor, and biased features.

'right to be forgotten' for all clients, allowing them to request data deletion from the

[Context and motivation] Large language models (LLMs) show notable results in natural language processing (NLP) tasks for requirements engineering (RE). However, their use is compromised by high computational cost, data sharing risks, and dependence on external services. In contrast, small language models (SLMs) offer a lightweight, locally deployable alternative. [Question/problem] It remains unclear how well SLMs perform compared to LLMs in RE tasks in terms of accuracy. [Results] Our preliminary study compares eight models, including three LLMs and five SLMs, on requirements classification tasks using the PROMISE, PROMISE Reclass, and SecReq datasets. Our results show that although LLMs achieve an average F1 score of 2% higher than SLMs, this difference is not statistically significant. SLMs almost reach LLMs performance across all datasets and even outperform them in recall on the PROMISE Reclass dataset, despite being up to 300 times smaller. We also found that dataset characteristics play a more significant role in performance than model size. [Contribution] Our study contributes with evidence that SLMs are a valid alternative to LLMs for requirements classification, offering advantages in privacy, cost, and local deployability.

The advent of Federated Learning (FL) highlights the practical necessity for the'right to be forgotten' for all clients, allowing them to request data deletion from the machine learning model's service provider. This necessity has spurred a growing demand for Federated Unlearning (FU). Feature unlearning has gained considerable attention due to its applications in unlearning sensitive, backdoor, and biased features. Existing methods employ the influence function to achieve feature unlearning, which is impractical for FL as it necessitates the participation of other clients, if not all, in the unlearning process. Furthermore, current research lacks an evaluation of the effectiveness of feature unlearning. To address these limitations, we define feature sensitivity in evaluating feature unlearning according to Lipschitz continuity.

As scam compounds in Southeast Asia continue to drive massive campaigns targeting victims around the world, WIRED took a deeper look at how Elon Musk's satellite internet service provider Starlink is keeping many of those compounds in Myanmar online. Meanwhile, FTC complaints obtained by WIRED allege that an "OpenAI" job scam used Telegram to recruit workers in Bangladesh for months before the fraudsters suddenly disappeared. WIRED published the inside story of Russian tech executive Vladislav Klyushin, who--at Vladimir Putin's behest--was part of a notable US-Russia prisoner swap last summer after he was convicted and incarcerated in the US for insider trading that netted him 93 million. Earlier this week, TVs at the headquarters of the Department of Housing and Urban Development in Washington, DC, showed an apparently AI-generated video on loop of Donald Trump kissing Elon Musk's feet. The words "LONG LIVE THE REAL KING" were superimposed over the video.

We introduce the Overall Performance Index (OPI), an intrinsic metric to evaluate retrieval-augmented generation (RAG) mechanisms for applications involving deep-logic queries. OPI is computed as the harmonic mean of two key metrics: the Logical-Relation Correctness Ratio and the average of BERT embedding similarity scores between ground-truth and generated answers. We apply OPI to assess the performance of LangChain, a popular RAG tool, using a logical relations classifier fine-tuned from GPT-4o on the RAG-Dataset-12000 from Hugging Face. Our findings show a strong correlation between BERT embedding similarity scores and extrinsic evaluation scores. Among the commonly used retrievers, the cosine similarity retriever using BERT-based embeddings outperforms others, while the Euclidean distance-based retriever exhibits the weakest performance. Furthermore, we demonstrate that combining multiple retrievers, either algorithmically or by merging retrieved sentences, yields superior performance compared to using any single retriever alone.

This paper explores continuous-time and state-space optimal stopping problems from a reinforcement learning perspective. We begin by formulating the stopping problem using randomized stopping times, where the decision maker's control is represented by the probability of stopping within a given time--specifically, a bounded, non-decreasing, c\`adl\`ag control process. To encourage exploration and facilitate learning, we introduce a regularized version of the problem by penalizing it with the cumulative residual entropy of the randomized stopping time. The regularized problem takes the form of an (n+1)-dimensional degenerate singular stochastic control with finite-fuel. We address this through the dynamic programming principle, which enables us to identify the unique optimal exploratory strategy. For the specific case of a real option problem, we derive a semi-explicit solution to the regularized problem, allowing us to assess the impact of entropy regularization and analyze the vanishing entropy limit. Finally, we propose a reinforcement learning algorithm based on policy iteration. We show both policy improvement and policy convergence results for our proposed algorithm.