Dynamic Demonstrations Controller for In-Context Learning
This addresses a key bottleneck in in-context learning for NLP practitioners by optimizing demonstration usage, though it is incremental as it builds on existing ICL methods.
The paper tackles the problem that in-context learning performance does not always improve with more demonstrations, proposing a Dynamic Demonstrations Controller to adjust the number dynamically, resulting in a 4.6% relative improvement across various LLMs and datasets.
In-context learning (ICL) is a new paradigm for natural language processing (NLP), where a large language model (LLM) observes a small number of demonstrations and a test instance as its input, and directly makes predictions without updating model parameters. Previous studies have revealed that ICL is sensitive to the selection and the ordering of demonstrations. However, there are few studies regarding the impact of the demonstration number on the ICL performance within a limited input length of LLM, because it is commonly believed that the number of demonstrations is positively correlated with model performance. In this paper, we found this conclusion does not always hold true. Through pilot experiments, we discover that increasing the number of demonstrations does not necessarily lead to improved performance. Building upon this insight, we propose a Dynamic Demonstrations Controller (D$^2$Controller), which can improve the ICL performance by adjusting the number of demonstrations dynamically. The experimental results show that D$^2$Controller yields a 4.6% relative improvement on ten different sizes of LLMs across ten datasets. Moreover, we also extend our method to previous ICL models and achieve competitive results.