Researchers from Carnegie Mellon University and Google explored the concept of delaying model outputs in language models by adding fake tokens. This technique, called pause training, was found to improve performance on various tasks, including extractive question-answering and reasoning. The team also discovered the optimal number of tokens for each task and observed that decreasing…

Machine learning models are being used to predict UI information and improve app accessibility and testing. Currently, these models rely on costly and error-prone human-labeled datasets. While some elements can be guessed from visuals or metadata, programmatically tapping the UI element is the only certain way to determine if it is “tappable.”

Mind-body interventions like mindfulness-based stress reduction (MBSR) can enhance well-being by improving awareness and control of physiological and cognitive states. Researchers examined the impact of MBSR on long-term physiological changes and well-being. They measured respiration rate (RR) and self-reported psychological well-being, distress, and medical symptoms. The study aimed to confirm the hypothesis that lower baseline…

Google set a new world record for calculating the most digits of Pi using the y-cruncher program running on Google Cloud. While math.pi has a precision of 15 digits, the article explores using Ramanujan’s formula and the Chudnovsky algorithm to approximate Pi to higher precision. The limitations of floating point numbers and the accumulation of…

The winners of Streamlit’s LLM Hackathon have been announced for building the most interesting Clarifai projects.

Researchers from Google DeepMind and Stanford University have developed a technique called “Analogical Prompting” to enhance the reasoning abilities of language models. Traditional prompts and pre-defined examples often fall short in guiding models to solve complex reasoning tasks. Analogical Prompting leverages the generative capabilities of language models to self-generate contextually relevant exemplars for each problem,…

The trend of AI-powered virtual girlfriends is rapidly escalating in the US, but experts are alarmed by the potential increase in loneliness among young men. Liberty Vittert, a data science professor, expressed concerns about the impact of AI girlfriends on male solitude. Multiple apps offering virtual girlfriend experiences have gained popularity, with some users developing…

The BBC has blocked OpenAI’s ChatGPT bot and the Common Crawl bot from scraping its news and media content. The decision follows a trend of websites blocking AI bots from using their data to train AI models. The BBC plans to explore using generative AI in content creation and operations, but acknowledges the risks concerning…

Advancements in generative AI have led to the creation of hyper-realistic digital content known as deepfakes, raising concerns about misinformation and fraud. Researchers have developed methods such as watermarking to distinguish between authentic and AI-generated material. The study found a trade-off between evasion and spoofing errors in image watermarking, as well as vulnerabilities to spoofing…