Claude 3.7 Sonnet's thinking process can be viewed transparently, which Anthropic believes might lend to better results.
Read more: Anthropic’s new Claude model can extend its thinking