Saved in:
Bibliographic Details
Main Authors: Jiang, Ye, Wang, Taihang, Liu, Youzheng, Wang, Yimin, Xia, Yuhan, Long, Yunfei
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.10675
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915616266387456
author Jiang, Ye
Wang, Taihang
Liu, Youzheng
Wang, Yimin
Xia, Yuhan
Long, Yunfei
author_facet Jiang, Ye
Wang, Taihang
Liu, Youzheng
Wang, Yimin
Xia, Yuhan
Long, Yunfei
contents In-context learning (ICL) for text classification, which uses a few input-label demonstrations to describe a task, has demonstrated impressive performance on large language models (LLMs). However, the selection of in-context demonstrations plays a crucial role and can significantly affect LLMs' performance. Most existing demonstration selection methods primarily focus on semantic similarity between test inputs and demonstrations, often overlooking the importance of label distribution alignment. To address this limitation, we propose a two-stage demonstration selection method, TopK + Label Distribution Divergence (L2D), which leverages a fine-tuned BERT-like small language model (SLM) to generate label distributions and calculate their divergence for both test inputs and candidate demonstrations. This enables the selection of demonstrations that are not only semantically similar but also aligned in label distribution with the test input. Extensive experiments across seven text classification benchmarks show that our method consistently outperforms previous demonstration selection strategies. Further analysis reveals a positive correlation between the performance of LLMs and the accuracy of the underlying SLMs used for label distribution estimation.
format Preprint
id arxiv_https___arxiv_org_abs_2511_10675
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Learn to Select: Exploring Label Distribution Divergence for In-Context Demonstration Selection in Text Classification
Jiang, Ye
Wang, Taihang
Liu, Youzheng
Wang, Yimin
Xia, Yuhan
Long, Yunfei
Computation and Language
Artificial Intelligence
Information Retrieval
In-context learning (ICL) for text classification, which uses a few input-label demonstrations to describe a task, has demonstrated impressive performance on large language models (LLMs). However, the selection of in-context demonstrations plays a crucial role and can significantly affect LLMs' performance. Most existing demonstration selection methods primarily focus on semantic similarity between test inputs and demonstrations, often overlooking the importance of label distribution alignment. To address this limitation, we propose a two-stage demonstration selection method, TopK + Label Distribution Divergence (L2D), which leverages a fine-tuned BERT-like small language model (SLM) to generate label distributions and calculate their divergence for both test inputs and candidate demonstrations. This enables the selection of demonstrations that are not only semantically similar but also aligned in label distribution with the test input. Extensive experiments across seven text classification benchmarks show that our method consistently outperforms previous demonstration selection strategies. Further analysis reveals a positive correlation between the performance of LLMs and the accuracy of the underlying SLMs used for label distribution estimation.
title Learn to Select: Exploring Label Distribution Divergence for In-Context Demonstration Selection in Text Classification
topic Computation and Language
Artificial Intelligence
Information Retrieval
url https://arxiv.org/abs/2511.10675