What Makes Good Data for Alignment? A Comprehensive Study of Automatic Data Selection in Instruction Tuning
Published in ICLR 2024, 2024
We present Deita, a comprehensive study of automatic data selection in instruction tuning, investigating what makes good data for alignment.
Authors: Wei Liu, Weihao Zeng, Keqing He, Yong Jiang, Junxian He
Recommended citation: Wei Liu*, Weihao Zeng*, Keqing He, Yong Jiang, Junxian He. (2024). "What Makes Good Data for Alignment? A Comprehensive Study of Automatic Data Selection in Instruction Tuning." ICLR 2024.
Download Paper