Researchers from the University of Tokyo & Google Propose Zero-shot-CoT: a Single Zero-Shot Prompt That Elicits a Chain of Thought From Large Language Models Across a Variety of Reasoning Tasks

Expanding linguistic models have played a big role in recent advances in natural language processing (NLP). Large language models’ (LLMs) effectiveness is frequently attributed to few-shot or zero-shot learning. It may tackle a variety of problems by simply conditioning the models on a few examples or instructions defining the problem. The process of conditioning the language model is known as “prompting,” and the manual construction of prompts has become a hot topic in NLP.

In contrast to the exceptional performance …

google language language models languagetechnology large language models reasoning researchers university university of tokyo