Anthropic's latest interpretability research: a new microscope to understand Claude's internal mechanisms

Selected text

Language models like Claude aren't programmed directly by humans—instead, they‘re trained on large amounts of data. During that training process, they learn their own strategies to solve problems. These strategies are encoded in the billions of computations a model performs for every word it writes. They arrive inscrutable to us, the model’s developers. This means that we don’t understand how models do most of the things they do.

like it says on the tin.

What if you assemble leaders from the biggest AI tech companies and gave them the mandate to transform higher education? What could possibly go wrong?

Claude Sonnet 4.5 is the best coding model in the world, strongest model for building complex agents, and best model at using computers.