Nebius

Repetitive employee queries can overwhelm HR departments and reduce time for strategic tasks. At Nebius, we implemented an AI Assistant…

Every year, people around the world spend up to 800 billion person-hours cleaning. It’s one of the most disliked chores, and businesses…

If you think New Year celebrations started on January 1, it depends on how you look at it — because in the AI world, the real fireworks…

LLM hallucinations are model outputs that appear ungrounded, inconsistent or nonsensical — a real bane for any AI developer. It’s quite…

We’ve built an open-source Kubernetes operator that runs and manages Slurm clusters as K8s resources. From this article, you’ll find out…

By uncovering surprising links between seemingly unrelated LLM architectures, a way might be paved for effective idea exchange and boosting…

It’s a little quieter now — as is often the case in August — in terms of big AI conferences, announcements, and product launches. Everyone…

The rapid evolution of deep learning models has brought about unprecedented growth in both their size and complexity. This trend, while…

Mixture of Experts (MoE) has become popular as an efficiency-boosting architectural component for LLMs. In this blog post, we’ll explore…

How well can LLMs abstract rules and how to test such ability? Here’s a research by the London Institute for Mathematical Sciences (LIMS).

Severe GPU scarcity and struggles with MLOps are forcing ML engineers to more and more divert focus from model development. Such a shift…

Why are datasets for LLMs so challenging? As with any machine learning task, data is half the battle (the other half being model efficiency…

In the next installment of our series of deep technical articles on AI research, let’s switch our attention to the famous LoRA, a low-rank…

Scaling your machine learning workloads will eventually require resource orchestration. Among the multiple solutions available, the most…

Retrieval-augmented generation (RAG) is a technique that enhances language models by combining generative AI with a retrieval component…

With this article, we are starting a new category on our blog, the one dedicated to AI research. Expect these posts to be very technical…

There are various aspects to optimize when training large models. It often lasts weeks and involves managing billions of rows of data, with…

The global network of ML engineers is divided into two parts: industrial and academic. The flow of information, methods of interaction, and…