Hallucinations in large language model (LLM) outputs severely limit their reliability in knowledge-intensive tasks such as question answering. To address this challenge, we introduce REFIND (Retrieval-augmented Factuality hallucINation Detection), a novel framework that detects hallucinated spans within LLM outputs by directly leveraging retrieved documents. As part of the REFIND, we propose the Context Sensitivity Ratio (CSR), a novel metric that quantifies the sensitivity of LLM outputs to retrieved evidence. This innovative approach enables REFIND to efficiently and accurately detect hallucinations, setting it apart from existing methods. In the evaluation, REFIND demonstrated robustness across nine languages, including low-resource settings, and significantly outperformed baseline models, achieving superior IoU scores in identifying hallucinated spans. This work highlights the effectiveness of quantifying context sensitivity for hallucination detection, thereby paving the way for more reliable and trustworthy LLM applications across diverse languages.

Multi-aspect controllable text generation aims to control text generation in attributes from multiple aspects, making it a complex but powerful task in natural language processing. Supervised fine-tuning methods are often employed for this task due to their simplicity and effectiveness. However, they still have some limitations: low rank adaptation (LoRA) only fine-tunes a few parameters and has suboptimal control effects, while full fine-tuning (FFT) requires significant computational resources and is susceptible to overfitting, particularly when data is limited. Moreover, existing works typically train multi-aspect controllable text generation models using only single-aspect annotated data, which results in discrepancies in data distribution; at the same time, accurately generating text with specific attributes is a challenge that requires strong attribute-aware capabilities. To address these limitations, we propose a lightweight, adaptive and attribute-aware framework for multi-aspect controllable text generation. Our framework can dynamically adjust model parameters according to different aspects of data to achieve controllable text generation, aiming to optimize performance across multiple aspects. Experimental results show that our framework outperforms other strong baselines, achieves state-of-the-art performance, adapts well to data discrepancies, and is more accurate in attribute perception.

We solve a regularized weighted low-rank approximation problem by a stochastic gradient descent on a manifold. To guarantee the convergence of our stochastic gradient descent, we establish a convergence theorem on manifolds for retraction-based stochastic gradient descents admitting confinements. On sample data from the Netflix Prize training dataset, our algorithm outperforms the existing stochastic gradient descent on Euclidean spaces. We also compare the accelerated line search on this manifold to the existing accelerated line search on Euclidean spaces.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. Why did humans start speaking? Scientists suggest genetics played a big role – and they say the evolution of this singular ability was key to our survival. A new study links a particular gene to the ancient origins of spoken language, proposing that a protein variant found only in humans may have helped us communicate in a novel way.

It's been nearly four years since Spotify first announced that lossless audio streaming was on tap, and here we are, still waiting. Now there's another burst of chatter about the long-delayed feature, with Bloomberg reporting that Spotify HiFi is still in the pipeline--for those willing to pay extra. According to Bloomberg, lossless Spotify streaming will come in a new add-on called Music Pro, which will offer other benefits including AI-powered music remixing tools as well as discounted concert tickets. And yes, Spotify Music Pro will cost extra--possibly as much as 5.99 on top of your current Spotify Premium subscription, Bloomberg reports. That means individual Spotify Premium subscribers could be paying roughly 18 a month for lossless music streaming privileges.

Spotify is rolling out a Music Pro tier later this year that will give users access to higher-quality audio and remixing tools, according to Bloomberg. The tier will reportedly cost users 6 per month on top of their 12 Premium subscription, but they'll be priced differently across regions and will be cheaper in less-developed markets. Many long-time Spotify subscribers, however, will probably say that they'll believe it when they see it. The service teased a high-fidelity streaming option way back in 2017 and had confirmed that it was working to provide users with access to lossless audio in 2021. Several reports about the feature's availability had come out over the years after the company's confirmation. In 2024, Bloomberg also reported that HiFi streaming is expected to arrive before the year ended as a 5 add-on.

Infamous for his expletive-laden viral rants at livestreamed awards shows, Fares is a refreshingly firy and unpredictable voice in an all too corporate industry. As he puts it, "It doesn't matter where I work or what I do, I will always say what I want. People say to me that that's refreshing – but isn't it weird that you cannot say what you think in interviews? Do we live in a fucking communist country? Obviously, you have got to respect certain boundaries, but to not even be able to express what you think personally about stuff? Yet while gamers know him as a grinning chaos merchant and passionate ambassador of co-op gameplay, in Fares' adopted homeland of Sweden, he is best known as an award-winning film director. Jalla! was a domestic box office success, while his 2005 drama Zozo was a more introspective work about his childhood experience of fleeing the Lebanese civil war. Twenty years, five feature films and three video games later, Zozo was just one of many cathartic endeavours for Fares. "I've always been a storyteller," he says. "When I was young, I'd draw my own comics.

SHOPPING – Contains affiliated content. Products featured in this Shopping Finder article are selected by our shopping writers. If you make a purchase using links on this page, Dailymail.co.uk will earn an affiliate commission. The day Apple fans have been waiting for is nearly here. After many months of rumours, the tech giant is finally due to unveil a slew of new products on Wednesday.

Oscar Wilde said, "The difference between literature and journalism is that journalism is unreadable, and literature is not read." Unfortunately, The digitally archived journalism of Oscar Wilde's 19th century often has no or poor quality Optical Character Recognition (OCR), reducing the accessibility of these archives and making them unreadable both figuratively and literally. This paper helps address the issue by performing OCR on "The Nineteenth Century Serials Edition" (NCSE), an 84k-page collection of 19th-century English newspapers and periodicals, using Pixtral 12B, a pre-trained image-to-text language model. The OCR capability of Pixtral was compared to 4 other OCR approaches, achieving a median character error rate of 1%, 5x lower than the next best model. The resulting NCSE v2.0 dataset features improved article identification, high-quality OCR, and text classified into four types and seventeen topics. The dataset contains 1.4 million entries, and 321 million words. Example use cases demonstrate analysis of topic similarity, readability, and event tracking. NCSE v2.0 is freely available to encourage historical and sociological research. As a result, 21st-century readers can now share Oscar Wilde's disappointment with 19th-century journalistic standards, reading the unreadable from the comfort of their own computers.

AI systems often exhibit political bias, influencing users' opinions and decision-making. While political neutrality-defined as the absence of bias-is often seen as an ideal solution for fairness and safety, this position paper argues that true political neutrality is neither feasible nor universally desirable due to its subjective nature and the biases inherent in AI training data, algorithms, and user interactions. However, inspired by Joseph Raz's philosophical insight that "neutrality [...] can be a matter of degree" (Raz, 1986), we argue that striving for some neutrality remains essential for promoting balanced AI interactions and mitigating user manipulation. Therefore, we use the term "approximation" of political neutrality to shift the focus from unattainable absolutes to achievable, practical proxies. We propose eight techniques for approximating neutrality across three levels of conceptualizing AI, examining their trade-offs and implementation strategies. In addition, we explore two concrete applications of these approximations to illustrate their practicality. Finally, we assess our framework on current large language models (LLMs) at the output level, providing a demonstration of how it can be evaluated. This work seeks to advance nuanced discussions of political neutrality in AI and promote the development of responsible, aligned language models.