Unveiling Selection Biases: Exploring Order and Token Sensitivity in Large Language Models

TL;DR

This paper investigates how order and token biases affect large language models' decision-making, quantifies their impact, and proposes mitigation strategies to improve robustness in selection tasks.

Contribution

It precisely quantifies selection biases related to order and token sensitivity, and develops mitigation strategies to improve LLM robustness.

Findings

Order and token biases significantly influence LLM decisions

Mitigation strategies reduce sensitivity and improve stability

Analysis across models and tasks informs better LLM design

Abstract

In this paper, we investigate the phenomena of "selection biases" in Large Language Models (LLMs), focusing on problems where models are tasked with choosing the optimal option from an ordered sequence. We delve into biases related to option order and token usage, which significantly impact LLMs' decision-making processes. We also quantify the impact of these biases through an extensive empirical analysis across multiple models and tasks. Furthermore, we propose mitigation strategies to enhance model performance. Our key contributions are threefold: 1) Precisely quantifying the influence of option order and token on LLMs, 2) Developing strategies to mitigate the impact of token and order sensitivity to enhance robustness, and 3) Offering a detailed analysis of sensitivity across models and tasks, which informs the creation of more stable and reliable LLM applications for selection problems.