Axios reported on Friday that Apple CEO Tim Cook will donate 1 million to Donald Trump's inauguration, the latest tech company figure to do so. These donations--which aren't covered by campaign finance law and can be unlimited--signal a clear willingness to work with the Trump administration and a desire to curry favor with the once and future president. The tech sector is particularly eager to suck up. In December, Meta and Amazon donated 1 million to the inauguration committee, and OpenAI's Sam Altman said he planned to do the same. Uber and its CEO Dara Khosrowshahi have both donated 1 million apiece (even as the company's chief legal officer Tony West is vice president Kamala Harris' brother-in-law).

Fox News Correspondent, William La Jeunesse, joins'Fox News Sunday' to discuss the evolution of A.I. and the push lawmakers are making to regulate it. ADAPT: The Trump administration's recent announcement of a sweeping deregulatory agenda for generative artificial intelligence (Gen AI) has created ripples across industries. This policy shift has implications for professionals and businesses alike, signaling a future where Gen AI development will accelerate quickly. If you want your work and business to survive this new acceleration, you need to adapt quickly to our increasingly disrupted environment. Zachary Levi attends the UK premiere of Shazam!

Microsoft has published a lengthy piece talking about its vision for artificial intelligence development over the next four years, under the incoming Trump administration. In the piece, the company has revealed that it's spending a total of 80 billion on AI-enabled data centers in 2025. Microsoft said it's building out the data centers to be able to train and deploy AI models, as well as to power its cloud-based applications. While that's the entirety of its budget for projects around the world, more than half of it will go towards building data centers in the United States. The company explained that none of the progress on AI the industry has achieved thus far would be possible "without new partnerships founded on large-scale infrastructure investments." It's now calling for the incoming Trump administration to expand the government's support for the advancement of AI within the US, such as providing the National Science Foundation and US universities more funding for research.

It may be a new year, but the hacks, scams, and dangerous people lurking online haven't gone anywhere. Just a day before the ball dropped, the United States Treasury Department said it had been hacked. Officials believe the attackers are an as-yet-unidentified Advanced Persistent Threat group linked to China's government that exploited flaws in remote tech support software made by BeyondTrust to carry out what the Treasury Department described as a "major" breach. The company told the Treasury on December 8 that the attackers stole an authentication key, which ultimately allowed them to access department computers. While the Treasury says the attackers were only able to steal "certain unclassified documents," new details have already begun to emerge, which we'll get into more below.

There's no doubt Great Britain lays claim to some of the greatest scientific discoveries and inventions that have changed the face of modern society. Now, MailOnline's interactive map reveals the birthplace of 30 of these famous British marvels, from stainless steel to the jet engine and the electric motor. Who can forget Alan Turing's Bombe machine, used to break Enigma-enciphered messages about enemy military operations during WWII? Turing developed the Bombe in 1939 at Bletchley Park in Buckinghamshire and hundreds were built, marking a crucial contribution to the war effort. Also on the map is the hovercraft invented by Christopher Cockerell in 1955 and first launched four years later on the the Isle of Wight.

Low-altitude mobility, exemplified by unmanned aerial vehicles (UAVs), has introduced transformative advancements across various domains, like transportation, logistics, and agriculture. Leveraging flexible perspectives and rapid maneuverability, UAVs extend traditional systems' perception and action capabilities, garnering widespread attention from academia and industry. However, current UAV operations primarily depend on human control, with only limited autonomy in simple scenarios, and lack the intelligence and adaptability needed for more complex environments and tasks. The emergence of large language models (LLMs) demonstrates remarkable problem-solving and generalization capabilities, offering a promising pathway for advancing UAV intelligence. This paper explores the integration of LLMs and UAVs, beginning with an overview of UAV systems' fundamental components and functionalities, followed by an overview of the state-of-the-art in LLM technology. Subsequently, it systematically highlights the multimodal data resources available for UAVs, which provide critical support for training and evaluation. Furthermore, it categorizes and analyzes key tasks and application scenarios where UAVs and LLMs converge. Finally, a reference roadmap towards agentic UAVs is proposed, aiming to enable UAVs to achieve agentic intelligence through autonomous perception, memory, reasoning, and tool utilization. Related resources are available at https://github.com/Hub-Tian/UAVs_Meet_LLMs.

\emph{Abrupt learning} is commonly observed in neural networks, where long plateaus in network performance are followed by rapid convergence to a desirable solution. Yet, despite its common occurrence, the complex interplay of task, network architecture, and learning rule has made it difficult to understand the underlying mechanisms. Here, we introduce a minimal dynamical system trained on a delayed-activation task and demonstrate analytically how even a one-dimensional system can exhibit abrupt learning through ghost points rather than bifurcations. Through our toy model, we show that the emergence of a ghost point destabilizes learning dynamics. We identify a critical learning rate that prevents learning through two distinct loss landscape features: a no-learning zone and an oscillatory minimum. Testing these predictions in recurrent neural networks (RNNs), we confirm that ghost points precede abrupt learning and accompany the destabilization of learning. We demonstrate two complementary remedies: lowering the model output confidence prevents the network from getting stuck in no-learning zones, while increasing trainable ranks beyond task requirements (\textit{i.e.}, adding sloppy parameters) provides more stable learning trajectories. Our model reveals a bifurcation-free mechanism for abrupt learning and illustrates the importance of both deliberate uncertainty and redundancy in stabilizing learning dynamics.

Large Language Models (LLMs) have been shown to exhibit various biases and stereotypes in their generated content. While extensive research has investigated bias in LLMs, prior work has predominantly focused on explicit bias, leaving the more nuanced implicit biases largely unexplored. This paper presents a systematic framework grounded in social psychology theories to investigate and compare explicit and implicit biases in LLMs. We propose a novel "self-reflection" based evaluation framework that operates in two phases: first measuring implicit bias through simulated psychological assessment methods, then evaluating explicit bias by prompting LLMs to analyze their own generated content. Through extensive experiments on state-of-the-art LLMs across multiple social dimensions, we demonstrate that LLMs exhibit a substantial inconsistency between explicit and implicit biases, where explicit biases manifest as mild stereotypes while implicit biases show strong stereotypes. Furthermore, we investigate the underlying factors contributing to this explicit-implicit bias inconsistency. Our experiments examine the effects of training data scale, model parameters, and alignment techniques. Results indicate that while explicit bias diminishes with increased training data and model size, implicit bias exhibits a contrasting upward trend. Notably, contemporary alignment methods (e.g., RLHF, DPO) effectively suppress explicit bias but show limited efficacy in mitigating implicit bias. These findings suggest that while scaling up models and alignment training can address explicit bias, the challenge of implicit bias requires novel approaches beyond current methodologies.

Dataset pruning aims to select a subset of a dataset for efficient model training. While data efficiency in natural language processing has primarily focused on within-corpus scenarios during model pre-training, efficient dataset pruning for task-specific fine-tuning across diverse datasets remains challenging due to variability in dataset sizes, data distributions, class imbalance and label spaces. Current cross-dataset pruning techniques for fine-tuning often rely on computationally expensive sample ranking processes, typically requiring full dataset training or reference models. We address this gap by proposing Swift Cross-Dataset Pruning (SCDP). Specifically, our approach uses TF-IDF embeddings with geometric median to rapidly evaluate sample importance. We then apply dataset size-adaptive pruning to ensure diversity: for smaller datasets, we retain samples far from the geometric median, while for larger ones, we employ distance-based stratified pruning. Experimental results on six diverse datasets demonstrate the effectiveness of our method, spanning various tasks and scales while significantly reducing computational resources. Source code is available at: https://github.com/he-y/NLP-Dataset-Pruning

Large Language Models (LLMs) are susceptible to generating harmful content when prompted with carefully crafted inputs, a vulnerability known as LLM jailbreaking. As LLMs become more powerful, studying jailbreak methods is critical to enhancing security and aligning models with human values. Traditionally, jailbreak techniques have relied on suffix addition or prompt templates, but these methods suffer from limited attack diversity. This paper introduces DiffusionAttacker, an end-to-end generative approach for jailbreak rewriting inspired by diffusion models. Our method employs a sequence-to-sequence (seq2seq) text diffusion model as a generator, conditioning on the original prompt and guiding the denoising process with a novel attack loss. Unlike previous approaches that use autoregressive LLMs to generate jailbreak prompts, which limit the modification of already generated tokens and restrict the rewriting space, DiffusionAttacker utilizes a seq2seq diffusion model, allowing more flexible token modifications. This approach preserves the semantic content of the original prompt while producing harmful content. Additionally, we leverage the Gumbel-Softmax technique to make the sampling process from the diffusion model's output distribution differentiable, eliminating the need for iterative token search. Extensive experiments on Advbench and Harmbench demonstrate that DiffusionAttacker outperforms previous methods across various evaluation metrics, including attack success rate (ASR), fluency, and diversity.