Interpretability

Goodfire launched probes that monitor AI agents from inside the model, aiming to reduce safety costs while catching risky behavior before it escalates.

Anthropic published research on translating Claude's internal representations into human-readable text.