Chain-of-Thought Scaling: A New Paper Changes the Calculus
Researchers demonstrate that step-by-step reasoning scales more predictably than token-level next-token prediction — with significant implications for test-time compute.
6 min
Frontier Model Benchmark Sweep: Where Things Stand in 2025
A comparative look at how the latest frontier models perform across reasoning, coding, and multimodal tasks.
8 min
Latest
Enterprise AI Adoption: Q1 2025 State of the Market
Fortune 500 deployments accelerated sharply in Q1. We break down what is actually shipping versus proof-of-concept, and which verticals are leading.
7 minConstitutional AI and the Interpretability Gap
Alignment techniques are advancing, but our ability to verify internal model behavior lags. A look at where interpretability research needs to go next.
10 minModel Context Protocol: Six Months In
MCP has shipped integrations across dozens of developer tools. We review what works, what does not, and where the spec should evolve.
9 min
Stay current
The LLM briefing, weekly.
One email. The most important model releases, research findings, and industry moves — curated from llms.blog.