UniICL: An Efficient Unified Framework Unifying Compression, Selection, and Generation

TL;DR

UniICL introduces a unified framework that combines demonstration compression, selection, and response generation to improve the efficiency and effectiveness of in-context learning in large language models.

Contribution

It proposes a novel unified approach that integrates demonstration compression, selection, and generation, with a caching strategy to enhance inference efficiency.

Findings

Outperforms existing methods in effectiveness on out-of-domain tasks.

Reduces computational cost through demonstration caching.

Maintains high performance with compressed demonstrations.

Abstract

In-context learning (ICL) enhances the reasoning abilities of Large Language Models (LLMs) by prepending a few demonstrations. It motivates researchers to introduce more examples to provide additional contextual information for the generation. However, existing methods show a significant limitation due to the problem of excessive growth in context length, which causes a large hardware burden. In addition, shallow-relevant examples selected by off-the-shelf tools hinder LLMs from capturing useful contextual information for generation. In this paper, we propose \textbf{UniICL}, a novel \textbf{Uni}fied \textbf{ICL} framework that unifies demonstration compression, demonstration selection, and final response generation. Furthermore, to boost inference efficiency, we design a tailored compression strategy that allows UniICL to cache compression results into \textbf{Demonstration Bank} (\textbf{DB}), which avoids repeated compression of the same demonstration. Extensive out-of-domain evaluations prove the advantages of UniICL in both effectiveness and efficiency.