Reasoning

Qwen3-Max-Thinking is Alibaba’s new major reasoning model. It not only measures parameters, but it also changes the way inference is done, with explicit control over the depth of thinking and …

Can a fully sovereign open logic model match state-of-the-art systems when every part of its training pipeline is transparent? Researchers at Mohammed Bin Zayed University of Artificial Intelligence (MBZUAI) have …

Estimate scaling has become one of the most effective methods to improve the quality and accuracy of answers in deployed LLMs. The idea is simple. If we are willing to …

Liquid AI has released LFM2.5-1.2b-Thinking, a 1.2 billion parameter reasoning model that runs entirely on device and fits in about 900 MB on a modern phone. What was required in …

Technology Innovation Institute (TII), Abu Dhabi, has released the Falcon-H1R-7B, a 7B parameter reasoning specialized model that matches or exceeds many 14B to 47B reasoning models in math, code and …

Liquid AI has introduced LFM2-2.6b-XP, an experimental checkpoint of its LFM2-2.6b language model trained with pure reinforcement learning on top of the existing LFM2 stack. The goal is simple, to …

Can a 3B model provide the logic of a 30B range by fixing the training recipe rather than the scaling parameters? The NanBiz LLM lab at Boss Zipin has released …

In this tutorial, we build an advanced meta-cognitive control agent that learns how to control the depth of its thinking. We treat reasoning as a spectrum, ranging from fast guesses …

Nvidia used this year’s Neurips conference to reveal new AI that it hopes will help accelerate progress toward widespread self-driving vehicles. At the event in San Diego, the company introduced …

How do you get GPT-5-level logic on real long-context, tool-use workloads without paying the quadratic attention and GPU costs that typically make those systems impractical? DeepSeek Research Introduces deepseek-v3.2 and …