When Long Helps Short: How Context Length in Supervised Fine-tuning Affects Behavior of Large Language Models.

Yingming Zheng, Hanqi Li, Kai Yu, Lu Chen

Browse the full EMNLP paper archive.