Flip-Flop Consistency: Unsupervised Training for Robustness to Prompt Perturbations in LLMs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hejabi, Parsa, Rahmati, Elnaz, Ziabari, Alireza S., Dehghani, Morteza
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915557251481600
author Hejabi, Parsa
Rahmati, Elnaz
Ziabari, Alireza S.
Dehghani, Morteza
author_facet Hejabi, Parsa
Rahmati, Elnaz
Ziabari, Alireza S.
Dehghani, Morteza
contents Large Language Models (LLMs) often produce inconsistent answers when faced with different phrasings of the same prompt. In this paper, we propose Flip-Flop Consistency ($F^2C$), an unsupervised training method that improves robustness to such perturbations. $F^2C$ is composed of two key components. The first, Consensus Cross-Entropy (CCE), uses a majority vote across prompt variations to create a hard pseudo-label. The second is a representation alignment loss that pulls lower-confidence and non-majority predictors toward the consensus established by high-confidence, majority-voting variations. We evaluate our method on 11 datasets spanning four NLP tasks, with 4-15 prompt variations per dataset. On average, $F^2C$ raises observed agreement by 11.62%, improves mean $F_1$ by 8.94%, and reduces performance variance across formats by 3.29%. In out-of-domain evaluations, $F^2C$ generalizes effectively, increasing $\overline{F_1}$ and agreement while decreasing variance across most source-target pairs. Finally, when trained on only a subset of prompt perturbations and evaluated on held-out formats, $F^2C$ consistently improves both performance and agreement while reducing variance. These findings highlight $F^2C$ as an effective unsupervised method for enhancing LLM consistency, performance, and generalization under prompt perturbations. Code is available at https://github.com/ParsaHejabi/Flip-Flop-Consistency-Unsupervised-Training-for-Robustness-to-Prompt-Perturbations-in-LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2510_14242
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Flip-Flop Consistency: Unsupervised Training for Robustness to Prompt Perturbations in LLMs
Hejabi, Parsa
Rahmati, Elnaz
Ziabari, Alireza S.
Dehghani, Morteza
Computation and Language
Machine Learning
Large Language Models (LLMs) often produce inconsistent answers when faced with different phrasings of the same prompt. In this paper, we propose Flip-Flop Consistency ($F^2C$), an unsupervised training method that improves robustness to such perturbations. $F^2C$ is composed of two key components. The first, Consensus Cross-Entropy (CCE), uses a majority vote across prompt variations to create a hard pseudo-label. The second is a representation alignment loss that pulls lower-confidence and non-majority predictors toward the consensus established by high-confidence, majority-voting variations. We evaluate our method on 11 datasets spanning four NLP tasks, with 4-15 prompt variations per dataset. On average, $F^2C$ raises observed agreement by 11.62%, improves mean $F_1$ by 8.94%, and reduces performance variance across formats by 3.29%. In out-of-domain evaluations, $F^2C$ generalizes effectively, increasing $\overline{F_1}$ and agreement while decreasing variance across most source-target pairs. Finally, when trained on only a subset of prompt perturbations and evaluated on held-out formats, $F^2C$ consistently improves both performance and agreement while reducing variance. These findings highlight $F^2C$ as an effective unsupervised method for enhancing LLM consistency, performance, and generalization under prompt perturbations. Code is available at https://github.com/ParsaHejabi/Flip-Flop-Consistency-Unsupervised-Training-for-Robustness-to-Prompt-Perturbations-in-LLMs.
title Flip-Flop Consistency: Unsupervised Training for Robustness to Prompt Perturbations in LLMs
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2510.14242