Despite the great potential of large language models(LLMs) in machine comprehension, it is still disturbing to fully count on them in real-world scenarios. This is probably because there is no rational explanation for whether the comprehension process of LLMs is aligned with that of experts. In this paper, we propose SCOP to carefully examine how LLMs perform during the comprehension process from a cognitive view. Specifically, it is equipped with a systematical definition of five requisite skills during the comprehension process, a strict framework to construct testing data for these skills, and a detailed analysis of advanced open-sourced and closed-sourced LLMs using the testing data. With SCOP, we find that it is still challenging for LLMs to perform an expert-level comprehension process. Even so, we notice that LLMs share some similarities with experts, e.g., performing better at comprehending local information than global information. Further analysis reveals that LLMs can be somewhat unreliable -- they might reach correct answers through flawed comprehension processes. Based on SCOP, we suggest that one direction for improving LLMs is to focus more on the comprehension process, ensuring all comprehension skills are thoroughly developed during training.

Video shows the moment drones started falling from the sky during a drone show at Eola Lake in Orlando, Florida on Dec. 21, 2024. A child was hospitalized on Saturday after being hit by a drone that was part of an Orlando, Florida holiday drone show. According to the Orlando Fire Department, a 7-year-old boy was transported to the hospital because of injuries sustained from the falling drones, FOX 35 in Orlando reported. In a video posted online by X user MosquitoCoFl, hundreds of drones being used as part of an aerial light show appeared to be flying into position before several started falling from the sky before slamming to the ground. A man could be heard saying to children nearby, "Oh no! I don't believe they're supposed to be falling."

Populism is a concept that is often used but notoriously difficult to measure. Common qualitative measurements like holistic grading or content analysis require great amounts of time and labour, making it difficult to quickly scope out which politicians should be classified as populist and which should not, while quantitative methods show mixed results when it comes to classifying populist rhetoric. In this paper, we develop a pipeline to train and validate an automated classification model to estimate the use of populist language. We train models based on sentences that were identified as populist and pluralist in 300 US governors' speeches from 2010 to 2018 and in 45 speeches of presidential candidates in 2016. We find that these models classify most speeches correctly, including 84% of governor speeches and 89% of presidential speeches. These results extend to different time periods (with 92% accuracy on more recent American governors), different amounts of data (with as few as 70 training sentences per category achieving similar results), and when classifying politicians instead of individual speeches. This pipeline is thus an effective tool that can optimise the systematic and swift classification of the use of populist language in politicians' speeches.

A Florida Christian school teacher was arrested this week after allegedly creating child sexual abuse materials using photos from the school yearbook and artificial intelligence (AI), according to authorities. The Pasco County Sheriff'sOffice said 67-year-old Steven Houser of New Port Richey faces charges for possession of child pornography. Deputies initiated an investigation after receiving an unspecified tip about Houser. Steven Guy Houser, a third-grade science teacher at a Christian school in New Port Richey, Florida, was allegedly found to be in possession of child pornography he created using yearbook photos and artificial intelligence. The investigation discovered that Beacon, a third-grade science teacher at Beacon Christian Academy, allegedly possessed two photos and three videos depicting child pornography.

Diffusion models have achieved great success in modeling continuous data modalities such as images, audio, and video, but have seen limited use in discrete domains such as language. Recent attempts to adapt diffusion to language have presented diffusion as an alternative to existing pretrained language models. We view diffusion and existing language models as complementary. We demonstrate that encoder-decoder language models can be utilized to efficiently learn high-quality language autoencoders. We then demonstrate that continuous diffusion models can be learned in the latent space of the language autoencoder, enabling us to sample continuous latent representations that can be decoded into natural language with the pretrained decoder. We validate the effectiveness of our approach for unconditional, class-conditional, and sequence-to-sequence language generation. We demonstrate across multiple diverse data sets that our latent language diffusion models are significantly more effective than previous diffusion language models.

It's never been about safety as much as it has control, serving and protecting only to the benefit of the status quo. In More than a Glitch, data journalist and New York University Associate Professor of Journalism Dr. Meredith Broussard, explores how and why we thought automating aspects of already racially-skewed legal, banking, and social systems would be a good idea. From facial recognition tech that doesn't work on dark-skinned folks to mortgage approval algorithms that don't work for dark-skinned folks, Broussard points to a dishearteningly broad array of initiatives that done more harm than good, regardless of their intention. In the excerpt below, Dr. Broussard looks at America's technochauavnistic history of predictive policing. Reprinted with permission from The MIT Press.

Knowledge graphs represent factual knowledge about the world as relationships between concepts and are critical for intelligent decision making in enterprise applications. New knowledge is inferred from the existing facts in the knowledge graphs by encoding the concepts and relations into low-dimensional feature vector representations. The most effective representations for this task, called Knowledge Graph Embeddings (KGE), are learned through neural network architectures. Due to their impressive predictive performance, they are increasingly used in high-impact domains like healthcare, finance and education. However, are the black-box KGE models adversarially robust for use in domains with high stakes? This thesis argues that state-of-the-art KGE models are vulnerable to data poisoning attacks, that is, their predictive performance can be degraded by systematically crafted perturbations to the training knowledge graph. To support this argument, two novel data poisoning attacks are proposed that craft input deletions or additions at training time to subvert the learned model's performance at inference time. These adversarial attacks target the task of predicting the missing facts in knowledge graphs using KGE models, and the evaluation shows that the simpler attacks are competitive with or outperform the computationally expensive ones. The thesis contributions not only highlight and provide an opportunity to fix the security vulnerabilities of KGE models, but also help to understand the black-box predictive behaviour of KGE models.

Having spent more than 25 years working with industry partners to educate and prepare the future workforce, it is not surprising to see that Florida has experienced growth in the technology sector. Across the nation, the U.S. Bureau of Labor Statistics estimates that computer and information technology occupations are projected to grow 11 percent from 2019 to 2029, much faster than the average for all occupations. Additionally, demand for skilled professionals in robotics and artificial intelligence is growing. The World Economic Forum estimates that while 85 million jobs will be displaced, 97 million new jobs will be created across 26 countries by 2025 due to the growth of artificial intelligence technology. From my conversations with industry leaders to the research and data I've studied, all signs lead me to believe that robotics and artificial intelligence will be a significant economic driver, surpassing the impact of Christopher Columbus' exploration of the New World in 1492.