Deep Tide TechFlow update: On July 25, University of Pennsylvania professor Ethan Mollick found that Claude may have stopped showing complete summary-level reasoning traces, and released before-and-after comparisons. Mollick believes this is a major loss that will harm a model’s interpretability. He noted that even summarized reasoning traces can help users diagnose errors in ways that other approaches can’t, and that these traces themselves have insight value. As of now, Anthropic’s official team has not provided any explanation for this change.