A few months before the 2025 International Mathematical Olympiad (IMO) in July, a three-person team at OpenAI made a long bet ...

OpenAI researchers reveal how their experimental model, devoid of any external aids, powered through hours-long proofs to earn a gold-medal score at the International Math Olympiad—and they discuss th ...

Google is releasing the high-performing Deep Think AI to select researchers, supporting advanced reasoning tests and future ...

If OpenAI's new model can solve grade-school math, it could pave the way for more powerful systems. This story is from The Algorithm, our weekly newsletter on AI. To get stories like this in your ...

A University of Queensland researcher has developed a new mathematical model to explain the evolution of the universe which, for the first time, includes collapsing regions of matter and expanding ...

Google released its first publicly available "multi-agent" AI system, which uses more computational resources, but produces ...

Researchers fine-tuned the Mistral 7B model to create Orca-Math, an SLM that solved grade school math problems better, at 86.8 percent, than models like Llama 2, GPT-3.5, and Gemini Pro — even ...

AI large language models have been especially weak on math. There are now several papers from Google Deep Mind, Alibaba and other universities where AI large language models are at Math Olympiad ...

On the MATH benchmark of competition level math word problems, for example, Meta’s model posted a score of 73.8, compared to GPT-4o’s 76.6 and Claude 3.5 Sonnet’s 71.1.

The model is also trained to answer PhD-level questions in subjects ranging from astrophysics to organic chemistry. In math olympiad questions, the new model is 83.3% accurate, versus 13.4% for ...