Topic models are a popular tool for understanding text collections, but their evaluation has been a point of contention. Automated evaluation metrics such as coherence are often used, however, their validity has been questioned for neural topic models (NTMs) and can overlook the benefits of a model in real world applications. To this end, we conduct the first evaluation of neural, supervised and classical topic models in an interactive task based setting. We combine topic models with a classifier and test their ability to help humans conduct content analysis and document annotation. From simulated, real user and expert pilot studies, the Contextual Neural Topic Model does the best on cluster evaluation metrics and human evaluations; however, LDA is competitive with two other NTMs under our simulated experiment and user study results, contrary to what coherence scores suggest. We show that current automated metrics do not provide a complete picture of topic modeling capabilities, but the right choice of NTMs can be better than classical models on practical tasks.

This report introduces LLaMandement, a state-of-the-art Large Language Model, fine-tuned by the French government and designed to enhance the efficiency and efficacy of processing parliamentary sessions (including the production of bench memoranda and documents required for interministerial meetings) by generating neutral summaries of legislative proposals. Addressing the administrative challenges of manually processing a growing volume of legislative amendments, LLaMandement stands as a significant legal technological milestone, providing a solution that exceeds the scalability of traditional human efforts while matching the robustness of a specialized legal drafter. We release all our fine-tuned models and training data to the community.

In response to rising concerns surrounding the safety, security, and trustworthiness of Generative AI (GenAI) models, practitioners and regulators alike have pointed to AI red-teaming as a key component of their strategies for identifying and mitigating these risks. However, despite AI red-teaming's central role in policy discussions and corporate messaging, significant questions remain about what precisely it means, what role it can play in regulation, and how precisely it relates to conventional red-teaming practices as originally conceived in the field of cybersecurity. In this work, we identify recent cases of red-teaming activities in the AI industry and conduct an extensive survey of the relevant research literature to characterize the scope, structure, and criteria for AI red-teaming practices. Our analysis reveals that prior methods and practices of AI red-teaming diverge along several axes, including the purpose of the activity (which is often vague), the artifact under evaluation, the setting in which the activity is conducted (e.g., actors, resources, and methods), and the resulting decisions it informs (e.g., reporting, disclosure, and mitigation). In light of our findings, we argue that while red-teaming may be a valuable big-tent idea for characterizing a broad set of activities and attitudes aimed at improving the behavior of GenAI models, gestures towards red-teaming as a panacea for every possible risk verge on security theater. To move toward a more robust toolbox of evaluations for generative AI, we synthesize our recommendations into a question bank meant to guide and scaffold future AI red-teaming practices.

Explanations of AI systems rarely address the information needs of people affected by algorithmic decision-making (ADM). This gap between conveyed information and information that matters to affected stakeholders can impede understanding and adherence to regulatory frameworks such as the AI Act. To address this gap, we present the "XAI Novice Question Bank": A catalog of affected stakeholders' information needs in two ADM use cases (employment prediction and health monitoring), covering the categories data, system context, system usage, and system specifications. Information needs were gathered in an interview study where participants received explanations in response to their inquiries. Participants further reported their understanding and decision confidence, showing that while confidence tended to increase after receiving explanations, participants also met understanding challenges, such as being unable to tell why their understanding felt incomplete. Explanations further influenced participants' perceptions of the systems' risks and benefits, which they confirmed or changed depending on the use case. When risks were perceived as high, participants expressed particular interest in explanations about intention, such as why and to what end a system was put in place. With this work, we aim to support the inclusion of affected stakeholders into explainability by contributing an overview of information and challenges relevant to them when deciding on the adoption of ADM systems. We close by summarizing our findings in a list of six key implications that inform the design of future explanations for affected stakeholder audiences.

The detection of toxic language in the Arabic language has emerged as an active area of research in recent years, and reviewing the existing datasets employed for training the developed solutions has become a pressing need. This paper offers a comprehensive survey of Arabic datasets focused on online toxic language. We systematically gathered a total of 54 available datasets and their corresponding papers and conducted a thorough analysis, considering 18 criteria across four primary dimensions: availability details, content, annotation process, and reusability. This analysis enabled us to identify existing gaps and make recommendations for future research works. For the convenience of the research community, the list of the analysed datasets is maintained in a GitHub repository (https://github.com/Imene1/Arabic-toxic-language).

The design of functional materials with desired properties is essential in driving technological advances in areas like energy storage, catalysis, and carbon capture. Generative models provide a new paradigm for materials design by directly generating entirely novel materials given desired property constraints. Despite recent progress, current generative models have low success rate in proposing stable crystals, or can only satisfy a very limited set of property constraints. Here, we present MatterGen, a model that generates stable, diverse inorganic materials across the periodic table and can further be fine-tuned to steer the generation towards a broad range of property constraints. To enable this, we introduce a new diffusion-based generative process that produces crystalline structures by gradually refining atom types, coordinates, and the periodic lattice. We further introduce adapter modules to enable fine-tuning towards any given property constraints with a labeled dataset. Compared to prior generative models, structures produced by MatterGen are more than twice as likely to be novel and stable, and more than 15 times closer to the local energy minimum. After fine-tuning, MatterGen successfully generates stable, novel materials with desired chemistry, symmetry, as well as mechanical, electronic and magnetic properties. Finally, we demonstrate multi-property materials design capabilities by proposing structures that have both high magnetic density and a chemical composition with low supply-chain risk. We believe that the quality of generated materials and the breadth of MatterGen's capabilities represent a major advancement towards creating a universal generative model for materials design.

Three American troops were killed and dozens more were injured in northeast Jordan Sunday in an attack that marked a major escalation of tensions in the region. The Islamic Resistance in Iraq, a loose coalition of Iran-backed militant groups, is claiming responsibility for the deadly attack. Per an analysis from the Pro-Israeli Washington Institute for Near East Policy, the "Islamic Resistance in Iraq," is not a singular unit per se but rather, an umbrella term used to tie the operations of various Iran-backed proxies in Iraq and Syria. The report determined that an umbrella term obscures responsibility, making it more difficult to determine who is exactly responsible for attacks on U.S. targets. IRAN-BACKED MILITIA KILLS 3 US TROOPS JUST WEEKS AFTER BIDEN SAID TEHRAN KNOWS'NOT TO DO ANYTHING' This satellite photo from Planet Labs PBC shows a military base known as Tower 22 in northeastern Jordan, on Oct. 12, 2023.

Iran, also under pressure at home, has held back from its own strikes on Israeli or American sites in retaliation for the assassination of its senior Revolutionary Guard commanders, which it blames on Israel. Earlier this month, in its first direct reply, it focused its fire on what was regarded as a "soft target" when it hit what it called a base of Israel's Mossad agency in Iraqi Kurdistan.

The U.S. and Iraq held an initial round of formal talks about ending the U.S.-led military mission in the country to fight against the Islamic State. Iraqi Prime Minister Mohammed Shia al-Sudani announced Sunday that he had sponsored "the commencement of the first round of bilateral dialogue between Iraq and the United States of America to end the mission of the Coalition in Iraq," according to a report from The Associated Press. That statement was followed by one from the coalition, which said military officials will assess "the threat of Daesh (IS), operational and environmental requirements and Iraqi Security Force capabilities" and a higher military commission will "work to set the conditions to transition the mission in Iraq," according to the report. U.S. soldiers train at al-Asad air base in western Iraq. While the initial talks come as U.S. forces have been under increased attacks in the region, including a drone attack in Jordan Sunday that killed three U.S. service members and injured 25 more, U.S. officials say that plans to end the mission in Iraq were first discussed last year and that the timing of the talks with Iraq were not related to the increased attacks.

Critics took aim at President Biden's Middle East policy after three Americans service members were killed in an attack on a base in Jordan near the border with Syria. "Hit Iran now," Sen. Lindsey Graham, R-S.C., said in a statement after the Sunday attack. Graham's comment comes after three U.S. service members were killed and 25 more were injured in a drone attack on northeast Jordan that sits close to the border with Syria, U.S. Central Command (CENTCOM) confirmed. "On Jan. 28, three U.S. service members were killed and 25 injured from a one-way attack UAS that impacted at a base in northeast Jordan, near the Syria border. As a matter of respect for the families and in accordance with DoD policy, the identities of the servicemembers will be withheld until 24 hours after their next of kin have been notified," CENTCOM said in a statement. "Updates will be provided as they become available."