# The Few-shot Dilemma: Over-prompting Large Language Models

Source: [The Few-shot Dilemma: Over-prompting Large Language Models](<https://www.sota2.com/research/paper/68fa7b661e209131cea88313>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-07-29T05:00:05.450Z

Authors: Yongjian Tang, Doruk Tuncel, Christian Koerner, Thomas Runkler

Published: 2025-09-16T16:00:06.000Z

## Abstract

Over-prompting, a phenomenon where excessive examples in prompts lead to diminished performance in Large Language Models (LLMs), challenges the conventional wisdom about in-context few-shot learning. To investigate this few-shot dilemma, we outline a prompting framework that leverages three standard few-shot selection methods - random sampling, semantic embedding, and TF-IDF vectors - and evaluate these methods across multiple LLMs, including GPT-4o, GPT-3.5-turbo, DeepSeek-V3, Gemma-3, LLaMA-3.1, LLaMA-3.2, and Mistral. Our experimental results reveal that incorporating excessive domain-specific examples into prompts can paradoxically degrade performance in certain LLMs, which contradicts the prior empirical conclusion that more relevant few-shot examples universally benefit LLMs. Given the trend of LLM-assisted software engineering and requirement analysis, we experiment with two real-world software requirement classification datasets. By gradually increasing the number of TF-IDF-selected and stratified few-shot examples, we identify their optimal quantity for each LLM. This combined approach achieves superior performance with fewer examples, avoiding the over-prompting problem, thus surpassing the state-of-the-art by 1% in classifying functional and non-functional requirements.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [FR/NFR Classification on PROMISE original (10-fold cross val)](<https://www.sota2.com/research/sota/fr-nfr-classification-on-promise-original-10-fold-cross-val.md>): FR/NFR Classification; PROMISE original (10-fold cross val); FR Precision: 94.

- [Multi-class classification on PROMISE original (10-fold cross-validation)](<https://www.sota2.com/research/sota/multi-class-classification-on-promise-original-10-fold-cross-validation.md>): Multi-class classification; PROMISE original (10-fold cross-validation); F1 Score (A): 100.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2509.13196>)

[PDF](<https://arxiv.org/pdf/2509.13196.pdf>)
