Meta AI Introduces DeepConf: First AI Method to Achieve 99.9% on AIME 2025 with Open-Source Models Using GPT-OSS-120B

Large language models (LLMs) have reshaped AI reasoning, with parallel thinking and self-consistency methods often cited as pivotal advances. However, these techniques face a fundamental trade-off: sampling multiple reasoning paths boosts accuracy but at a steep computational cost. A team of researchers from Meta AI and UCSD introduce Deep Think with Confidence (DeepConf), a new … Read more

How Google’s AI can help transform health professions education

Acknowledgements The research described here is a joint effort across Google Research, Google for Health, Google DeepMind, and partnering teams. The following researchers contributed to this work: Kevin McKee, Dan Gillick, Irina Jurenka, Markus Kunesch, Kaiz Alarakyia, Miriam Schneider, Jenn Sturgeon, Maggie Shiels, Amy Wang, Roma Ruparel, Anna Iurchenko, Mahvish Nagda, Julie Anne Séguin, Divya … Read more

Commerce Department Blocks Natcast’s CHIPS Funding

The U.S. Commerce Department says it will not abide by an agreement to fund the U.S. CHIPS and Science Act’s R&D through the nonprofit set up to administer the program, called Natcast. Instead, it handed operational control to the National Institute of Standards and Technology (NIST). Natcast was created in 2023 to oversee the National … Read more

[2508.19807] Bootstrapping Learned Cost Models with Synthetic SQL Queries

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. … Read more

Australia’s Large Language Model Landscape: Technical Assessment

Key Points No flagship, globally competitive, locally developed LLM (such as GPT-4, Claude 3.5, LLaMA 3.1) has yet emerged from Australia. Australian research and commerce currently rely primarily on international LLMs, which are frequently used but have measurable limitations on Australian English and cultural context. Kangaroo LLM is the only major open-source, locally developed LLM … Read more