Exploring the Robustness of In-Context Learning with Noisy Labels

Recently, the mysterious In-Context Learning (ICL) ability exhibited by Transformer architectures, especially in large language models (LLMs), has sparked significant research interest. However, the resilience of Transformers' in-context learning capabilities in the presence of noisy samples, prevalent in both training corpora and prompt demonstrations, remains underexplored. In this paper, inspired by prior research that studies ICL ability using simple function classes, we take a closer look at this problem by investigating the robustness of Transformers against noisy labels. Furthermore, we delve deeper into this problem by exploring whether introducing noise into the training set, akin to a form of data augmentation, enhances such robustness during inference, and find that such noise can indeed improve the robustness of ICL. Overall, our fruitful analysis and findings provide a comprehensive understanding of the resilience of Transformer models against label noises during ICL and provide valuable insights into the research on Transformers in natural language processing. In recent years, Large Language Models (LLMs) have achieved significant success across various tasks in real-world applications. Transformer (Vaswani et al., 2017), as the typical backbone architecture, also emerges an intriguing ability known as In-Context Learning (ICL) (Brown et al., 2020; Dong et al., 2023a) that the model can learn a new task with only a few input-output pairs demonstrated during inference without modifying any model parameters.