AAAI 2026technical0 citations

Identifying and Analyzing Performance-Critical Tokens in Large Language Models

Yu Bai, Heyan Huang, Cesare Spinoso-Di Piano, Sanxing Chen, Marc-Antoine Rondeau, Yang Gao, Jackie Chi Kit Cheung

Abstract

In-context learning (ICL) has emerged as an effective solution for few-shot learning with large language models (LLMs). However, how LLMs leverage demonstrations to specify a task and learn a corresponding computational function through ICL is underexplored. Drawing from the way humans learn from content-label mappings in demonstrations, we categorize the tokens in an ICL prompt into content, stopword, and template tokens. Our goal is to identify the types of tokens whose representations directly influence LLM

BibTeX
@inproceedings{aaai2026_identifyingandan,
  title = {Identifying and Analyzing Performance-Critical Tokens in Large Language Models},
  author = {Yu Bai and Heyan Huang and Cesare Spinoso-Di Piano and Sanxing Chen and Marc-Antoine Rondeau and Yang Gao and Jackie Chi Kit Cheung},
  booktitle = {AAAI 2026},
  year = {2026}
}