? DeepSeek-R1-Lite-Preview is now live: unleashing supercharged reasoning power! ? o1-preview-level performance on AIME & MATH benchmarks. ? Transparent thought process in real-time.
In the passkey retrieval task, which requires finding hidden numbers within 1 million tokens of irrelevant text, the model achieves 100 percent accuracy regardless of the information's position in the ...
Nvidia achieved new record numbers in the third quarter of fiscal year 2025, with revenue rising to $35.1 billion. According to CEO Jensen Huang, the company is just at the beginning of two ...
A commission of the US Congress recommends a massive program to develop artificial general intelligence in order to take the lead in competition with China. Other key technologies are also in the ...
The code provides for two central documents: the "Safety and Security Framework" (SSF) and the "Safety and Security Report" (SSR). The SSF is the overarching framework that sets out the basic risk ...
AI start-up Genmo has released its Mochi 1 video model as an open source version. According to the company, it is the largest publicly available AI model for video generation to date, with 10 billion ...
A new study by researchers at Bytedance Research and Tsinghua University shows that current AI video models like OpenAI's Sora can create impressive visuals but fail to understand the physical laws ...
Meta is expanding into business-focused AI services with a new division led by tech veteran Clara Shih, the company announced Thursday. The new Business AI group aims to make Meta's AI technology ...
Microsoft Research has created a new AI system called Magentic-One that can handle complex computer tasks by working with web content and files. The system uses multiple specialized AI agents instead ...
Mistral AI adds web search and image generation to its Le Chat AI assistant, while introducing a new visual model that performs well on industry benchmarks. Le Chat users can now access current web ...
A new research method reveals how AI systems learn concepts and are more capable than previously thought. Analysis in "concept space" reveals surprising results and offers clues for better training ...
In the MLPerf Training 4.1 benchmarks, the Nvidia Blackwell platform delivered 2.2 times more performance per GPU compared to Hopper in the LLM benchmark Llama 2 70B fine-tuning and 2 times more ...