The latest addition is the Phi-4 Reasoning — a 14 billion-parameter model built by applying a supervised fine-tuning (SFT) algorithm to the Phi-4 base model. The researchers also derived the Phi-4 ...

Many top language models now err on the side of caution, refusing harmless prompts that merely sound risky – an ‘over-refusal' behavior that affects their usefulness in real-world scenarios. A new ...

While usage of Artificial Intelligence apps in the workplace is surging, security and governance are being left behind.

Alibaba Group Holding Ltd. updated its video-generating model for the second time in one month as Chinese firms look to ...

Today, CPU performance bottlenecks on servers managing multi-modal and large language model workloads are a driving factor for low 30-40% average GPU utilization rates. This results in expensive ...

China’s advances in AI are allowing humanoid developers to pair the robots’ already impressive hardware with the software needed to make them economically valuable.

Imagine having a virtual assistant that is always productive, never sleeps, and consistently produces excellent outcomes. Supa provides just that—an intelligent, all-in-one AI assistant designed to ...

A new AI platform named Manus is turning heads in the global tech community for its ability to handle tasks traditionally ...

Rogue communication devices not listed in product documents have been found in some Chinese solar power inverters by U.S ...

New Poe data reveals major shifts in AI market share as OpenAI and Google gain ground while specialized reasoning models surge to 10% of usage in 2025.

The Qwen3 family of AI models can now be deployed via large language model platforms Ollama, LM Studio, SGLang and vLLM.