notesum.ai

Published at December 5

cs.CL

cs.AI

Released Date: December 5, 2024

Authors: Changcheng Li¹, Xiangyu Wang¹, Qiuju Chen¹, Xiren Zhou¹, Huanhuan Chen¹

Aff.: ¹University of Science and Technology of China

Method	GPQA	TruthfulQA	GSM8K \bigstrut
GPT-4o mini	38.8%	55.4%	93.6% \bigstrut[t]
+ CoT	39.1%	57.6%	93.3%
+ 3-shot	41.1%	58.9%	93.1%
+ CoT 1-shot	40.2%	57.3%	93.9%
+ MTMT	44.0%	58.5%	93.9% \bigstrut[b]