Large language models (LLMs) are powerful tools capable of handling diverse tasks. Comparing and selecting appropriate LLMs for specific tasks requires systematic evaluation methods, as models exhibit varying capabilities across different domains. However, finding suitable benchmarks is difficult given the many available options. This complexity not only increases the risk of benchmark misuse and misinterpretation but also demands substantial effort from LLM users, seeking the most suitable benchmarks for their specific needs. To address these issues, we introduce BenchmarkCards, an intuitive and validated documentation framework that standardizes critical benchmark attributes such as objectives, methodologies, data sources, and limitations. Through user studies involving benchmark creators and users, we show that BenchmarkCardscan simplify benchmark selection and enhance transparency, facilitating informed decision-making in evaluating LLMs.

The six-episode series brings The Great Dying to life. More information Adding us as a Preferred Source in Google by using this link indicates that you would like to see more of our content in Google News results. During The Great Dying, upwards of 80 percent of life on Earth was wiped out. Breakthroughs, discoveries, and DIY tips sent six days a week. By signing up, you confirm you are 16+, will receive newsletters and promotional content and agree to our Terms of Use and acknowledge the data practices in our Privacy Policy .

Large language models (LMs) often struggle to generate diverse, human-like creative content, raising concerns about the long-term homogenization of human thought through repeated exposure to similar outputs. Yet scalable methods for evaluating LM output diversity remain limited, especially beyond narrow tasks such as random number or name generation, or beyond repeated sampling from a single model. To address this gap, we introduce INFINITY-CHAT, a largescale dataset of 26K diverse, real-world, open-ended user queries that admit a wide range of plausible answers with no single ground truth. We introduce the first comprehensive taxonomy for characterizing the full spectrum of open-ended prompts posed to LMs, comprising 6 top-level categories (e.g., creative content generation, brainstorm & ideation) that further breaks down to 17 subcategories.

A Border Force officer and his handler who worked for Chinese intelligence in the UK have been jailed. Chi Leung Peter Wai, 40, was sentenced to 10 years and Chung Biu Bill Yuen, 65, given an eight year term after being found guilty of assisting a foreign intelligence service, an offence under the National Security Act. Wai used his position as a Border Force officer to access to the Home Office computer system to track Hong Kong dissidents in the UK, was also convicted of misconduct in public office. The judge Mrs Justice Cheema-Grubb told the men that their actions threaten the sovereignty of the state during sentencing remarks at the Old Bailey on Thursday. The dual Chinese-British nationals were found guilty after a trial last month.

Image classifiers are information-discarding machines, by design. Yet, how these models discard information remains mysterious. We hypothesize that one way for image classifiers to reach high accuracy is to zoom to the most discriminative region in the image and then extract features from there to predict image labels, discarding the rest of the image. Studying six popular networks ranging from AlexNet to CLIP, we find that proper framing of the input image can lead to the correct classification of 98.91% of ImageNet images. Furthermore, we uncover positional biases in various datasets, especially a strong center bias in two popular datasets: ImageNet-A and ObjectNet. Finally, leveraging our insights into the potential of zooming, we propose a test-time augmentation (TTA) technique that improves classification accuracy by forcing models to explicitly perform zoom-in operations before making predictions. Our method is more interpretable, accurate, and faster than MEMO, a state-of-the-art (SOTA) TTA method. We introduce ImageNet-Hard, a new benchmark that challenges SOTA classifiers including large vision-language models even when optimal zooming is allowed.

Large language models (LLMs) have emerged as a dominant AI paradigm due to their exceptional text understanding and generation capabilities. However, their tendency to generate inconsistent or erroneous outputs challenges their reliability, especially in high-stakes domains requiring accuracy and trustworthiness. Existing research primarily focuses on detecting and mitigating model misbehavior in general-purpose scenarios, often overlooking the potential of integrating domain-specific knowledge. In this work, we advance misbehavior detection by incorporating domain knowledge. The core idea is to design a general specification language that enables domain experts to customize domain-specific constraints in a lightweight and intuitive manner, supporting later runtime monitoring of LLM outputs.

This week only, every donation is doubled! Halfway through our Summer Membership Drive, we're still well behind where we need to be. But there's good news: This week, every donation will be doubled up, to $50,000 We need you right now. We need you right now. Trump's Justice Department Backs Elon Musk in Data Center Lawsuit DOJ urges judge to throw NAACP's legal action over xAI's gas turbines in Mississippi.

Recent advances have enabled LLM-powered AI agents to autonomously execute complex tasks by combining language model reasoning with tools, memory, and web access. But can these systems be trusted to follow deployment policies in realistic environments, especially under attack? To investigate, we ran the largest public red-teaming competition to date, targeting 22 frontier AI agents across 44 realistic deployment scenarios. Participants submitted 1.8 million promptinjection attacks, with over 60,000 successfully eliciting policy violations such as unauthorized data access, illicit financial actions, and regulatory noncompliance. We use these results to build the Agent Red Teaming (ART) benchmark--a curated set of high-impact attacks--and evaluate it across 19state-of-the-art models.

Apple CEO warns price rises'unavoidable' amid AI boom The prices of Apple products will have to increase due to the new demand for memory chips from the artificial intelligence boom, outgoing Apple CEO Tim Cook has told The Wall Street Journal. "Unfortunately, price increases are unavoidable," he told the newspaper on Wednesday, adding that his company has been "trying to shield customers from the increases" but that it had become "unsustainable." It is also unclear, for instance, how much the price of Apple's iPhone 18, which is expected to launch in September, will be affected. "There's less supply at a time when consumers want devices and the memory guys are passing along huge price increases," Cook said. Citing an estimate from research firm TechInsights, the Journal reported that Apple would need to increase the price of its iPhone Pro model by $270 to maintain its current profit margin .

We argue that the machine learning value chain is structurally unsustainable due to an economic data processing inequality: each state in the data cycle from inputs to model weights to synthetic outputs refines technical signal but strips economic equity from data generators. We show, by analyzing seventy-three public data deals, that the majority of value accrues to aggregators, with documented creator royalties rounding to zero and widespread opacity of deal terms. This is not just an economic welfare concern: as data and its derivatives become economic assets, the feedback loop that sustains current learning algorithms is at risk. We identify three structural faults--missing provenance, asymmetric bargaining power, and nondynamic pricing--as the operational machinery of this inequality. In our analysis, we trace these problems along the machine learning value chain and propose an Equitable Data-Value Exchange (EDVEX) Framework to enable a minimal market that benefits all participants. Finally, we outline research directions where our community can make concrete contributions to data deals and contextualize our position with related and orthogonal viewpoints.