Anthropic built a tool that reads Claude’s unspoken thoughts. Then it caught the model scheming

Anthropic has built something close to a mind-reading tool for its own AI. What it found sits somewhere between a breakthrough and an unsettling party trick. Anthropic researchers now have the clearest view yet of what a large language model does while it thinks. In a paper published on the company’s Transformer Circuits site, they […]
This story continues at The Next Web