Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
| dc.contributor.author | Wei, Jason | |
| dc.contributor.author | Wang, Xuezhi | |
| dc.contributor.author | Schuurmans, Dale | |
| dc.contributor.author | Bosma, Maarten | |
| dc.contributor.author | Ichter, Brian | |
| dc.contributor.author | Xia, Fei | |
| dc.contributor.author | Chi, Ed | |
| dc.contributor.author | Le, Quoc | |
| dc.contributor.author | Zhou, Denny | |
| dc.date.accessioned | 2025-06-02T13:09:56Z | |
| dc.date.available | 2025-06-02T13:09:56Z | |
| dc.date.issued | 2022-01-10 | |
| dc.description | Chain-of-Thought prompting 기법은 LLM의 다단계 추론 과정을 유도하여 복잡한 문제 해결력을 높이는 데 효과적임을 보입니다. ©2022 Google Research | |
| dc.description.abstract | We explore how generating a chain of thought -- a series of intermediate reasoning steps -- significantly improves the ability of large language models to perform complex reasoning. In particular, we show how such reasoning abilities emerge naturally in sufficiently large language models via a simple method called chain of thought prompting, where a few chain of thought demonstrations are provided as exemplars in prompting. Experiments on three large language models show that chain of thought prompting improves performance on a range of arithmetic, commonsense, and symbolic reasoning tasks. The empirical gains can be striking. For instance, prompting a 540B-parameter language model with just eight chain of thought exemplars achieves state of the art accuracy on the GSM8K benchmark of math word problems, surpassing even finetuned GPT-3 with a verifier. | |
| dc.description.sponsorship | Google Research | |
| dc.identifier.uri | https://arxiv.org/abs/2201.11903 | |
| dc.identifier.uri | http://data.inu.ac.kr/handle/123456789/1955 | |
| dc.language.iso | en_US | |
| dc.publisher | arXiv | |
| dc.subject | Chain-of-Thought | |
| dc.subject | Prompt Engineering | |
| dc.subject | Reasoning | |
| dc.subject | LLM | |
| dc.title | Chain-of-Thought Prompting Elicits Reasoning in Large Language Models | |
| dc.type | Article |