Coverage-based Example Selection for In-Context Learning
This addresses the issue of redundant or incomplete example selection for in-context learning, which is crucial for improving task performance across various LLMs and datasets, though it is incremental as it builds on existing metrics.
The paper tackled the problem of selecting informative examples for in-context learning in large language models, showing that BERTScore-Recall (BSR) and its set-level extension outperform standard similarity-based methods, with improvements of up to 17 points on average for compositional tasks.
In-context learning (ICL), the ability of large language models to perform novel tasks by conditioning on a prompt with a few task examples, requires these examples to be informative about the test instance. The standard approach of independently ranking and selecting the most similar examples selects redundant examples while omitting important information. In this work, we show that BERTScore-Recall (BSR) selects better examples that demonstrate more of the salient aspects, e.g. reasoning patterns, of the test input. We further extend BSR and many standard metrics to easily optimizable set-level metrics, giving still better coverage of those salient aspects. On 15 datasets spanning 6 tasks and with 7 diverse LLMs, we show that (1) BSR is the superior metric for in-context example selection across the board, and (2) for compositional tasks, set selection using Set-BSR outperforms independent ranking by up to 17 points on average and, despite being training-free, surpasses methods that leverage task or LLM-specific training.