APTO Releases Training Dataset to Enhance the Mathematical Reasoning Capabilities of Large Language Models (LLMs)

As generative AI use continues to increase, accuracy has become the most important metric and a key factor in decisions ...

AI21’s Jamba Reasoning 3B Redefines What “Small” Means in LLMs — 250K Context on a Laptop

AI21's new model Jamba Reasoning 3B can support a 250K or more token context window while running on an iPhone.

Samsung AI researcher's new, open reasoning model TRM outperforms models 10,000X larger — on specific problems

The trend of AI researchers developing new, small open source generative models that outperform far larger, proprietary peers continued this week with yet another staggering advancement. Alexia ...

Samsung AI Research Team Builds A Tiny Model With Big Power

Samsung’s AI lab in Montreal new Tiny Recursive Model with only 7M parameters performs as well, if not better in some ...

Fractal’s science-led proposal under IndiaAI Mission targets math, STEM, and medical reasoning

Fractal Analytics, part of IndiaAI Mission, aims to advance reasoning models with a focus on AI development in India.
EdSource

A greater role in math education for parents: mathematical reasoning at home

September 18, 2025 - Meet Allison Saiki, who teaches students how to manage money, pay rent and open retirement accounts, with a class currency she calls "Saiki Cents." While policymakers, researchers ...

AI that can think: The rise of reasoning models that can mimic human logic

AI reasoning models like OpenAI's o1 and o3-mini redefine machine thinking by analyzing, inferring, and connecting data ...
The Texas Observer

Reasoning Mind: How Big Oil and Texas Politicians Came to Love Math

At one point during this year’s SXSWedu—a slick Austin conference heavy on marketing for education technology—an audience member stood and asked a panel of Texas lawmakers the question most in the ...