Unveiling In-Context Learning: A Coordinate System to Understand Its Working Mechanism

TL;DR

This paper introduces a two-dimensional coordinate system to unify different theories of in-context learning in large language models, providing a systematic framework to understand its mechanisms across classification and generation tasks.

Contribution

It proposes a novel coordinate system that explains ICL behavior through perception and cognition variables, integrating conflicting views and enabling comprehensive analysis.

Findings

The coordinate system effectively explains ICL behavior across multiple tasks.

The peak inverse rank metric detects LLMs' task recognition ability.

Experiments reveal how similarity and task recognition influence ICL performance.

Abstract

Large language models (LLMs) exhibit remarkable in-context learning (ICL) capabilities. However, the underlying working mechanism of ICL remains poorly understood. Recent research presents two conflicting views on ICL: One emphasizes the impact of similar examples in the demonstrations, stressing the need for label correctness and more shots. The other attributes it to LLMs' inherent ability of task recognition, deeming label correctness and shot numbers of demonstrations as not crucial. In this work, we provide a Two-Dimensional Coordinate System that unifies both views into a systematic framework. The framework explains the behavior of ICL through two orthogonal variables: whether similar examples are presented in the demonstrations (perception) and whether LLMs can recognize the task (cognition). We propose the peak inverse rank metric to detect the task recognition ability of LLMs and study LLMs' reactions to different definitions of similarity. Based on these, we conduct extensive experiments to elucidate how ICL functions across each quadrant on multiple representative classification tasks. Finally, we extend our analyses to generation tasks, showing that our coordinate system can also be used to interpret ICL for generation tasks effectively.