Addressing Order Sensitivity of In-Context Demonstration Examples in Causal Language Models
Published in Findings of ACL 2024, 2024
We analyze how the ordering of in-context examples affects causal language model performance and propose strategies to reduce order sensitivity, improving robustness across downstream tasks.
Recommended citation: Yanzheng Xiang, Hanqi Yan, Lin Gui, Yulan He. 2024. "Addressing Order Sensitivity of In-Context Demonstration Examples in Causal Language Models." In Findings of ACL 2024.
Download Paper