For the last few years, chain-of-thought prompting has become the central method for reasoning in large language models. By encouraging models to “think aloud,” researchers found that step-by-step ...

In recent years, despite large models demonstrating excellent performance in natural language processing tasks, a significant ...

Proportional reasoning failures seem to constitute most errors in probabilistic reasoning, yet there is little empirical evidence about its role for attaining probabilistic knowledge and how to ...