Artificial intelligence models and research encompass the foundational algorithms, architectures, and scientific studies that drive system capabilities. Truncated covers new model releases, underlying technical advances, and published research papers.
AI-generated summary Nace published a technical report regarding Drex, which is a family of four open-weight decision models designed to return probabilities for permitted choices without generating word-by-word explanations. Nace reports that when Drex Max, a 27-billion-parameter model, states it is at least 93 percent sure, the company reports it is right 97 percent of the time.
AI-generated summary Model distillation transfers behavior from a larger or more capable teacher into a smaller student model using teacher outputs, internal signals, or softened probability targets. This process differs from simple compression because it requires teaching the smaller model rather than simply removing parameters.
AI-generated summary Nous Research fine-tuned Alibaba's Qwen3.8-27B on 53 million tokens to create Arete-27B and reduce Chinese state propaganda rates from 22 percent to under one percent. According to Nous, the resulting open-weight model maintains its general capabilities while dropping the propaganda rate and increasing factual accuracy.
AI-generated summary Devin labels the alias claude-donut-eap, which reverses in the Devin CLI parser to a model called Silicon. Silicon appears to be a new family with low, medium, high, xhigh, and max variants according to the source.
The top AI news in your email delivered each morning. Choose categories