Product Habits

Apple found cracks in LLMs' reasoning capabilities