Tags: Performance

At Computex, AMD unveiled its latest innovations, including the Radeon RX 9060 XT graphics cards and new Threadripper CPUs, marking a significant advancement in high-performance…

French AI startup Mistral has launched the Mistral Medium 3 AI model, designed for efficiency and high performance, particularly in coding and STEM applications. Priced…

This week, the AI community is focused on the launch of Olmo 2 1B, a 1-billion-parameter model from the nonprofit AI research institute Ai2, which…

Tokenization is a fundamental process in natural language processing (NLP) that involves breaking down text into smaller units called tokens, which can be words, subwords,…

On Wednesday, Microsoft introduced new AI models aimed at enhancing reasoning capabilities, competing with benchmarks like OpenAI’s o3-mini. The models include Phi 4 mini reasoning…

Salesforce has launched a groundbreaking AI research initiative to tackle “jagged intelligence,” which refers to the inconsistent performance of AI systems. This initiative aims to…

The Palmyra X5, a newly launched enterprise AI model, is poised to transform business automation and cost efficiency. Capable of processing up to 1,500 pages…

The integration of advanced models into applications is crucial in today’s tech landscape, enhancing search-augmented assistants, educational tools, and user support systems. These models improve…

OpenAI plans to launch its first open language model since GPT-2 in early summer, aiming to boost reasoning capabilities and compete with existing models. Led…

Training large language models (LLMs) on reasoning trajectories and tool use significantly boosts their effectiveness in complex, multi-step reasoning tasks. This approach enhances problem-solving capabilities,…