Browsing: Research

Most web agents today drive a browser one action at a time. The model receives the current page state — as a screenshot or DOM text…

Instruction-tuned language models refuse harmful requests. But which part of the model is actually responsible — and how does that mechanism get installed during training? A…

In 1996, Guinea-Bissau seemed like an ideal research post for budding pediatrician Lone Graff Stensballe. Her supervisor, a fellow Dane named Peter Aaby, had spent nearly…

AI research in 2025 was defined by major shifts. The industry moved beyond chatbots and into reasoning systems, autonomous agent and multimodal systems. Last year, companies…

Training large language models on long sequences has a well-known problem: attention is expensive. The scaled dot-product attention (SDPA) at the core of every transformer scales…

Pre-training large language models is expensive enough that even modest efficiency improvements can translate into meaningful cost and time savings. Nous Research is releasing Token Superposition…

Researchers at Tilde Research have released Aurora, a new optimizer for training neural networks that addresses a structural flaw in the widely-used Muon optimizer. The flaw…

Large language models are no longer just about scale. In 2026, the most important LLM research is focused on making models safer, more controllable, and more…

Your smartwatch knows your heart rate, your sleep schedule, and several other types of health data. But what happens when someone else gets access to all…

NotebookLM has been my go-to AI research tool for a while; something I rely on almost every day. In my line of work, research is the…