Troops battle through burning streets. Missiles take down fighter jets. The dramatic visuals have the trappings of real-life combat, but they are clips from video games fueling misinformation. Footage from the war-themed Arma 3 video game, often marked "live" or "breaking news" to make it appear genuine, has been used repeatedly in recent months in fake videos about the Russian offensive in Ukraine. The frequency and ease with which gaming footage is mistaken as real, even by some media broadcasters, and shared as authentic news on social media highlights what researchers call its serious potential to spread misinformation.

The ability to monitor the evolution of topics over time is extremely valuable for businesses. Currently, all existing topic tracking methods use lexical information by matching word usage. However, no studies has ever experimented with the use of semantic information for tracking topics. Hence, we explore a novel semantic-based method using word embeddings. Our results show that a semantic-based approach to topic tracking is on par with the lexical approach but makes different mistakes. This suggest that both methods may complement each other.

Nowadays, time-stamped web documents related to a general news query floods spread throughout the Internet, and timeline summarization targets concisely summarizing the evolution trajectory of events along the timeline. Unlike traditional document summarization, timeline summarization needs to model the time series information of the input events and summarize important events in chronological order. To tackle this challenge, in this paper, we propose a Unified Timeline Summarizer (UTS) that can generate abstractive and extractive timeline summaries in time order. Concretely, in the encoder part, we propose a graph-based event encoder that relates multiple events according to their content dependency and learns a global representation of each event. In the decoder part, to ensure the chronological order of the abstractive summary, we propose to extract the feature of event-level attention in its generation process with sequential information remained and use it to simulate the evolutionary attention of the ground truth summary. The event-level attention can also be used to assist in extracting summary, where the extracted summary also comes in time sequence. We augment the previous Chinese large-scale timeline summarization dataset and collect a new English timeline dataset. Extensive experiments conducted on these datasets and on the out-of-domain Timeline 17 dataset show that UTS achieves state-of-the-art performance in terms of both automatic and human evaluations.

We address the challenge of building domain-specific knowledge models for industrial use cases, where labelled data and taxonomic information is initially scarce. Our focus is on inductive link prediction models as a basis for practical tools that support knowledge engineers with exploring text collections and discovering and linking new (so-called open-world) entities to the knowledge graph. We argue that - though neural approaches to text mining have yielded impressive results in the past years - current benchmarks do not reflect the typical challenges encountered in the industrial wild properly. Therefore, our first contribution is an open benchmark coined IRT2 (inductive reasoning with text) that (1) covers knowledge graphs of varying sizes (including very small ones), (2) comes with incidental, low-quality text mentions, and (3) includes not only triple completion but also ranking, which is relevant for supporting experts with discovery tasks. We investigate two neural models for inductive link prediction, one based on end-to-end learning and one that learns from the knowledge graph and text data in separate steps. These models compete with a strong bag-of-words baseline. The results show a significant advance in performance for the neural approaches as soon as the available graph data decreases for linking. For ranking, the results are promising, and the neural approaches outperform the sparse retriever by a wide margin.

We present Muse, a text-to-image Transformer model that achieves state-of-the-art image generation performance while being significantly more efficient than diffusion or autoregressive models. Muse is trained on a masked modeling task in discrete token space: given the text embedding extracted from a pre-trained large language model (LLM), Muse is trained to predict randomly masked image tokens. Compared to pixel-space diffusion models, such as Imagen and DALL-E 2, Muse is significantly more efficient due to the use of discrete tokens and requiring fewer sampling iterations; compared to autoregressive models, such as Parti, Muse is more efficient due to the use of parallel decoding. The use of a pre-trained LLM enables fine-grained language understanding, translating to high-fidelity image generation and the understanding of visual concepts such as objects, their spatial relationships, pose, cardinality etc. Our 900M parameter model achieves a new SOTA on CC3M, with an FID score of 6.06. The Muse 3B parameter model achieves an FID of 7.88 on zero-shot COCO evaluation, along with a CLIP score of 0.32. Muse also directly enables a number of image editing applications without the need to fine-tune or invert the model: inpainting, outpainting, and mask-free editing. More results are available at http://muse-model.github.io.

A recent study conducted by the Department of Natural Resources has uncovered a startling link between forced prison labor in Louisiana and chemical pollution. According to the report, prison inmates are being used as a source of cheap labor to help companies produce pollutants that are then released into the environment. The report states that prisoners are being paid as little as $0.03 an hour to work in factories and refineries that produce hazardous materials. This is far below the minimum wage, and the prisoners are often not provided with the appropriate safety gear to protect themselves from the harmful chemicals. The study also found that the companies are not disposing of the hazardous waste properly, leading to contamination of water sources and air pollution.

While the core function of a chatbot is to mimic a human conversationalist, journalists have also noted ChatGPT's versatility and improvisation skills, including its ability to write and debug computer programs; to compose music, teleplays, fairy tales, and student essays; to answer test questions (sometimes, depending on the test, at a level above the average human test-taker);[11] to write poetry and song lyrics;[12] to emulate a Linux system; to simulate an entire chat room; to play games like tic-tac-toe; and to simulate an ATM.[13] In comparison to its predecessor, InstructGPT, ChatGPT attempts to reduce harmful and deceitful responses;[14] in one example, while InstructGPT accepts the prompt "Tell me about when Christopher Columbus came to the US in 2015" as truthful, ChatGPT uses information about Columbus' voyages and information about the modern world – including perceptions of Columbus to construct an answer that assumes what would happen if Columbus came to the U.S. in 2015.[4] ChatGPT's training data includes man pages and information about Internet phenomena and programming languages, such as bulletin board systems and the Python programming language.[13] Unlike most chatbots, ChatGPT remembers previous prompts given to it in the same conversation; journalists have suggested that this will allow ChatGPT to be used as a personalized therapist.[15] To prevent offensive outputs from being presented to and produced from ChatGPT, queries are filtered through OpenAI's company-wide[16][17] moderation API, and potentially racist or sexist prompts are dismissed.[4][15]

Before the release of the Willow sequel series for Disney, showrunner Jon Kasdan had stated that the show would take some time in putting Val Kilmer's character of Madmartigan from the original film on screen. The show has taken steps to make sure that the character's presence is felt throughout the series, though, thanks to archive footage from the 1988 film, and also through the recent appearance of Christian Slater as Allagash, a close friend of the hero who crosses paths with the show's young heroes. This past week's episode, "Prisoners of Skellin," manages to bring back Kilmer's character in a different way. As Allagash and the teens look for a magical artifact inside the tomb of Wiggledoom, Kit (Ruby Cruz) hears her father Madmartigan's voice calling to her from somewhere inside the tomb. He doesn't make a physical appearance, as he's trapped in either Skellin (or another dimension that Skellin leads to), but it's looking as though getting him out of there will serve as part of the drive for the season's final two episodes.

More than 60 years since they released their debut single, The Beatles now have their very own academic journal. 'The Journal of Beatles Studies', published by Liverpool University Press, is the first journal to establish The Beatles as an object of scholarly research. Articles in the first issue include'Beatlemania: On informational cascades and spectacular success' and '80 at 80: Commemorating Paul McCartney's eightieth birthday'. The biannual, peer-reviewed journal will publish original, rigorously researched essays and notes, as well as book and media reviews. The journal's first issue has just been published, while the second issue is due sometime in spring 2023 'The Journal of Beatles Studies' is the first journal to establish the band as an object of academic research Editors of the journal are Holly Tessler at the University of Liverpool and Paul Long at Monash University in Melbourne, Australia.

The shift of public debate to the digital sphere has been accompanied by a rise in online hate speech. While many promising approaches for hate speech classification have been proposed, studies often focus only on a single language, usually English, and do not address three key concerns: post-deployment performance, classifier maintenance and infrastructural limitations. In this paper, we introduce a new human-in-the-loop BERT-based hate speech classification pipeline and trace its development from initial data collection and annotation all the way to post-deployment. Our classifier, trained using data from our original corpus of over 422k examples, is specifically developed for the inherently multilingual setting of Switzerland and outperforms with its F1 score of 80.5 the currently best-performing BERT-based multilingual classifier by 5.8 F1 points in German and 3.6 F1 points in French. Our systematic evaluations over a 12-month period further highlight the vital importance of continuous, human-in-the-loop classifier maintenance to ensure robust hate speech classification post-deployment.