A team of Apple researchers has released a paper scrutinising the mathematical reasoning capabilities of large language models (LLMs), suggesting that while these models can exhibit abstract reasoning ...