The Power of Adaptation: Boosting In-Context Learning through Adaptive Prompting
Shuzhang Cai, Twumasi Mensah-Boateng, Xander Kuksov, Jing Yuan, Shaojie Tang
TL;DR
This work tackles the challenge of selecting informative exemplars for in-context learning in large language models, where fixed exemplars can underrepresent the task space and introduce redundancy. It introduces Adaptive-Prompt, an iterative, uncertainty-guided exemplar-selection framework that recomputes uncertainty after each annotation and adaptively expands the exemplar set $E$. Across arithmetic, commonsense, and symbolic reasoning tasks, Adaptive-Prompt generally outperforms non-adaptive baselines (e.g., Active-Prompt, Random-CoT, Auto-CoT) on GPT-3.5 Turbo and, to a lesser extent, on GPT-4o Mini, highlighting the value of adaptivity in prompting. The approach reduces redundancy and enhances informativeness in exemplars, with robust performance across annotators and settings, though gains can diminish for stronger or weaker models depending on task and prompt length.
Abstract
Large Language Models (LLMs) have demonstrated exceptional abilities across a broad range of language-related tasks, including generating solutions to complex reasoning problems. An effective technique to enhance LLM performance is in-context learning, which encourages a step-by-step reasoning process by including explanatory examples to guide the model's responses. However, selecting appropriate exemplars for the model poses a challenge, as each dataset demands a distinct set of exemplars to enable the LLM to learn effectively and perform well on the test set. Current studies often rely on uncertainty- or diversity-based selection strategies to select exemplars for annotation and to improve model learning. However, these studies typically employ a non-adaptive approach, selecting a set of exemplars all at once. We argue that this non-adaptive strategy may result in a set of exemplars with high redundancy in terms of the knowledge covered, ultimately reducing their overall informativeness. To address this limitation, we propose \textsc{Adaptive-Prompt}, a novel method that adaptively selects exemplars by leveraging model feedback from previously chosen exemplars. Experimental results show that \textsc{Adaptive-Prompt} significantly enhances LLM performance across a variety of reasoning tasks.