Since Google Gemini attributed part of its success in mathematics competitions to 'parallel thinking', how to enable large models to master the ability to explore multiple reasoning paths in parallel ...

The Parallel-R1 framework uses reinforcement learning to teach models how to explore multiple reasoning paths at once, ...

Today’s desktop computers are as powerful as supercomputers were 20 years ago, said computer science professor Joel Adams. And 20 years from now, the average desktop may be as powerful as the ...

Experiments show that Parallel-R1 not only brings an average accuracy improvement of up to 8.4% across multiple mathematical benchmarks but also achieves a 42.9% performance leap in the AIME25 test ...

Forty Form Two students of Shun Lee Catholic Secondary School could not hide their surprise at seeing more than 100 teachers waiting for them as they filed into a temporary classroom at the school ...