Search AI Products and News

Recently, the National Astronomical Observatories of China (NAOC) and Alibaba Cloud jointly announced the launch of Jinwu, the world's first solar large model, marking a significant step in the deep integration of solar physics research and artificial intelligence technology. The model, built on Alibaba Cloud's open-source Tongyi Qianwen framework, boasts a prediction accuracy exceeding 91% for M5-class solar flares, achieving the highest global level for this type of prediction. This achievement not only improves the accuracy of space weather forecasting but also provides new techniques for addressing the potential terrestrial impacts of solar activity.

On March 26th, at the Spring Summit of the New Energy Digital Asset Community in Suzhou, Ant Group officially launched EnergyTS, a time-series model for energy and power. This marks a significant breakthrough for Chinese companies in the field of energy AI. Specifically designed for the new energy sector, EnergyTS demonstrated significantly higher accuracy in photovoltaic power generation prediction compared to leading international general-purpose time-series models such as Google's (TimesFM-V2.0) and Amazon's (Chronos-Large). In T+1 day prediction, the model's mean absolute error is only...

Following previous attention in the speech AI field, OpenAI continues its exploration, releasing three new self-developed speech models: gpt-4o-transcribe, gpt-4o-mini-transcribe, and gpt-4o-mini-tts. The most notable is gpt-4o-transcribe. These new models are currently...

A German research team has developed and released FIORA, an open-source graph neural network (GNN) model designed to improve the accuracy of mass spectrometry analysis. This advancement is expected to significantly contribute to the progress of precision medicine.

Alibaba Tongyi Lab has open-sourced ViDoRAG, a novel visual document retrieval and question answering system. ViDoRAG boasts impressive accuracy, achieving a remarkable 79.4% in benchmark tests. This system represents a significant advancement in the field of computer vision and information retrieval.

ElevenLabs has unveiled its new Scribe speech-to-text model, boasting unprecedented accuracy levels, particularly in English where it achieves a remarkable 96.7% accuracy rate.

Crunchbase announced that it will abandon its role as a historical data provider and transform into an AI-driven prediction engine. This new platform will be able to predict startup funding rounds, acquisitions, and growth trajectories with an accuracy of up to 95%, marking a significant strategic shift for Crunchbase. Crunchbase CEO Jagger McConnell stated in an interview that the traditional historical data industry is in decline. He believes that data companies relying solely on historical data...

Social media giant Meta recently announced that they have developed a new device capable of reading brain neural signals to enable text input. This research achievement is detailed by Meta's scientists in two studies, utilizing advanced brain imaging technology and deep learning AI models to successfully decode the brain signals of individuals while they type, even reconstructing complete sentences. Specifically, this technology relies on a scanning device known as magnetoencephalography (MEG), which can capture the faint magnetic signals emitted by the brain.

Recently, Tencent Technology (Shenzhen) Co., Ltd. published a patent regarding a training method and related equipment for large language models on the Tianyancha app. The patent is titled 'Training Method, Device, Computer Equipment, and Storage Medium for Large Language Models' and aims to enhance the learning capacity and accuracy of large language models through innovative training methods. In the training process of large language models, traditional methods often rely on a single text summary, which may lead to model overfitting and negatively impact the accuracy and diversity of generated content. However, Tencent's new...

Recently, a study led by the Austrian Institute of Complexity Science (CSH) revealed that despite large language models (LLMs) excelling in various tasks, they show significant shortcomings when tackling advanced history questions. The research team tested three leading models, including OpenAI's GPT-4, Meta's Llama, and Google's Gemini, with disappointing results. Note: The image was generated by AI, and the image is licensed by Midjourney to evaluate these models in history.

Recently, researchers at Keele University developed a new AI tool capable of identifying fake news with an accuracy of up to 99%, providing crucial support in combating online misinformation. The research team includes Dr. Uchenna Anie from the School of Computer Science and Mathematics, Dr. Sangeeta Sangeeta, and Dr. Patricia Aso-Ayobode. The team employed various machine learning techniques to design a model that scans news content and assesses its credibility. The researchers used an 'ensemble voting' technique, which...

Recently, Meta has announced a groundbreaking research achievement by developing a new type of memory layer technology that can significantly enhance the factual accuracy of large language models (LLMs) and achieve unprecedented expansion in parameter scale. This technology not only challenges traditional neural network expansion methods but also provides new directions for future AI architecture design. The core of this research lies in utilizing a trainable key-value lookup mechanism to add extra parameters to the model without increasing computational load (FLOPs). This method's core thinking is...

The application of artificial intelligence in healthcare has reached a major breakthrough! A study conducted by several top institutions, including Harvard University and Stanford University, shows that OpenAI's o1-preview model demonstrates astonishing capabilities in multiple medical reasoning tasks, even surpassing human doctors. This research not only evaluated the model's performance on medical multiple-choice question benchmarks but also focused on its diagnostic and management capabilities in simulated real clinical scenarios, yielding impressive results. Researchers conducted five experiments to assess o1-