# Prompt Perturbation Consistency Learning for Robust Language Models

arXiv: https://arxiv.org/abs/2402.15833 · HTML: https://mlsanity.com/paper/2402.15833/

## Key facts

- Cited by 26 papers in the map (data through 14 Oct 2026).
- Builds on 31 of its 42 references that are in the map.
- First posted on arXiv 24 Feb 2024.

## Builds on (most cited first, 30 of 31)

- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?](https://mlsanity.com/paper/2202.12837/) (2022)
- [Gorilla: Large Language Model Connected with Massive APIs](https://mlsanity.com/paper/2305.15334/) (2023)
- [Unsupervised Data Augmentation for Consistency Training](https://mlsanity.com/paper/1904.12848/) (2019)
- [EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks](https://mlsanity.com/paper/1901.11196/) (2019)
- [Improving Neural Machine Translation Models with Monolingual Data](https://mlsanity.com/paper/1511.06709/) (2015)
- [Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment](https://mlsanity.com/paper/1907.11932/) (2019)
- [BERT-ATTACK: Adversarial Attack Against BERT Using BERT](https://mlsanity.com/paper/2004.09984/) (2020)
- [Is ChatGPT a General-Purpose Natural Language Processing Task Solver?](https://mlsanity.com/paper/2302.06476/) (2023)
- [A Survey of Data Augmentation Approaches for NLP](https://mlsanity.com/paper/2105.03075/) (2021)
- [Snips Voice Platform: an embedded Spoken Language Understanding system for private-by-design voice interfaces](https://mlsanity.com/paper/1805.10190/) (2018)
- [Do Prompt-Based Models Really Understand the Meaning of their Prompts?](https://mlsanity.com/paper/2109.01247/) (2021)
- [PromptRobust: Towards Evaluating the Robustness of Large Language Models on Adversarial Prompts](https://mlsanity.com/paper/2306.04528/) (2023)
- [BERT for Joint Intent Classification and Slot Filling](https://mlsanity.com/paper/1902.10909/) (2019)
- [The Unreliability of Explanations in Few-shot Prompting for Textual Reasoning](https://mlsanity.com/paper/2205.03401/) (2022)
- [MASSIVE: A 1M-Example Multilingual Natural Language Understanding Dataset with 51 Typologically-Diverse Languages](https://mlsanity.com/paper/2204.08582/) (2022)
- [An Empirical Survey of Data Augmentation for Limited Data Learning in NLP](https://mlsanity.com/paper/2106.07499/) (2021)
- [An Analysis of Simple Data Augmentation for Named Entity Recognition](https://mlsanity.com/paper/2010.11683/) (2020)
- [A Survey on Spoken Language Understanding: Recent Advances and New Frontiers](https://mlsanity.com/paper/2103.03095/) (2021)
- [MELM: Data Augmentation with Masked Entity Language Modeling for Low-Resource NER](https://mlsanity.com/paper/2108.13655/) (2021)
- [Distantly-Supervised Named Entity Recognition with Noise-Robust Learning and Language Model Augmented Self-Training](https://mlsanity.com/paper/2109.05003/) (2021)
- [On Robustness of Prompt-based Semantic Parsing with Large Pre-trained Language Model: An Empirical Study on Codex](https://mlsanity.com/paper/2301.12868/) (2023)
- [HiddenCut: Simple Data Augmentation for Natural Language Understanding with Better Generalization](https://mlsanity.com/paper/2106.00149/) (2021)
- [Data Augmentation for Spoken Language Understanding via Joint Variational Generation](https://mlsanity.com/paper/1809.02305/) (2018)
- [Controlled Text Generation for Data Augmentation in Intelligent Artificial Agents](https://mlsanity.com/paper/1910.03487/) (2019)
- [On Robustness of Neural Semantic Parsers](https://mlsanity.com/paper/2102.01563/) (2021)
- [Unsupervised Data Augmentation with Naive Augmentation and without Unlabeled Data](https://mlsanity.com/paper/2010.11966/) (2020)
- [DialogVCS: Robust Natural Language Understanding in Dialogue System Upgrade](https://mlsanity.com/paper/2305.14751/) (2023)

## Built on by (most cited first, 26 of 26)

- [Brittlebench: Quantifying LLM robustness via prompt sensitivity](https://mlsanity.com/paper/2603.13285/) (2026)
- [Robust Reward Modeling via Causal Rubrics](https://mlsanity.com/paper/2506.16507/) (2025)
- [Prompt Perturbations Reveal Human-Like Biases in Large Language Model Survey Responses](https://mlsanity.com/paper/2507.07188/) (2025)
- [Whose Facts Win? LLM Source Preferences under Knowledge Conflicts](https://mlsanity.com/paper/2601.03746/) (2026)
- [When Punctuation Matters: A Large-Scale Comparison of Prompt Robustness Methods for LLMs](https://mlsanity.com/paper/2508.11383/) (2025)
- [Are Small Language Models Ready to Compete with Large Language Models for Practical Applications?](https://mlsanity.com/paper/2406.11402/) (2024)
- [Evaluating the Robustness of Retrieval-Augmented Generation to Adversarial Evidence in the Health Domain](https://mlsanity.com/paper/2509.03787/) (2025)
- [Learning to Explore with Parameter-Space Noise: A Deep Dive into Parameter-Space Noise for Reinforcement Learning with Verifiable Rewards](https://mlsanity.com/paper/2602.02555/) (2026)
- [AbstRaL: Augmenting LLMs' Reasoning by Reinforcing Abstract Thinking](https://mlsanity.com/paper/2506.07751/) (2025)
- [Explainable Mapper: Charting LLM Embedding Spaces Using Perturbation-Based Explanation and Verification Agents](https://mlsanity.com/paper/2507.18607/) (2025)
- [Same Meaning, Different Scores: Lexical and Syntactic Sensitivity in LLM Evaluation](https://mlsanity.com/paper/2602.17316/) (2026)
- [Flip-Flop Consistency: Unsupervised Training for Robustness to Prompt Perturbations in LLMs](https://mlsanity.com/paper/2510.14242/) (2025)
- [FragileFlow: Spectral Control of Correct-but-Fragile Predictions for Foundation Model Robustness](https://mlsanity.com/paper/2605.08896/) (2026)
- [How Important is `Perfect' English for Machine Translation Prompts?](https://mlsanity.com/paper/2507.09509/) (2025)
- [Limited Preference Data? Learning Better Reward Model with Latent Space Synthesis](https://mlsanity.com/paper/2509.26074/) (2025)
- [R2V Agent: Teaching SLMs When to Ask for Help](https://mlsanity.com/paper/2605.16604/) (2026)
- [UpSkill: Mutual Information Skill Learning for Structured Response Diversity in LLMs](https://mlsanity.com/paper/2602.22296/) (2026)
- [Where Do Prompt Perturbations Break Generation? A Segment-Level View of Robustness in LoRA-Tuned Language Models](https://mlsanity.com/paper/2605.01605/) (2026)
- [Intended Target Identification for Anomia Patients with Gradient-based Selective Augmentation](https://mlsanity.com/paper/2506.14203/) (2025)
- [Are Humans as Brittle as Large Language Models?](https://mlsanity.com/paper/2509.07869/) (2025)
- [Harmonizing Diverse Models: A Layer-wise Merging Strategy for Consistent Generation](https://mlsanity.com/paper/2510.14915/) (2025)
- [Seeing the Needle in the Haystack: Towards Weakly-Supervised Log Instance Anomaly Localization via Counterfactual Perturbation](https://mlsanity.com/paper/2605.10988/) (2026)
- [Emergent retokenization symmetry in large language models: phenomenology and applications](https://mlsanity.com/paper/2606.15521/) (2026)
- [Prompt Perturbation for Reliable LLM Evaluation over Comparison Graphs](https://mlsanity.com/paper/2606.17634/) (2026)
- [Beyond Prompt Engineering: A Systematic Analysis of Prompt Lexical Sensitivity and Its Impacts on Quality](https://mlsanity.com/paper/2608.20349/) (2026)
- [Prompt-Robust Language Models: Which Training Strategies Work?](https://mlsanity.com/paper/2609.01217/) (2026)

## Abstract (arXiv)

Large language models (LLMs) have demonstrated impressive performance on a number of natural language processing tasks, such as question answering and text summarization. However, their performance on sequence labeling tasks such as intent classification and slot filling (IC-SF), which is a central component in personal assistant systems, lags significantly behind discriminative models. Furthermore, there is a lack of substantive research on the robustness of LLMs to various perturbations in the input prompts. The contributions of this paper are three-fold. First, we show that fine-tuning sufficiently large LLMs can produce IC-SF performance comparable to discriminative models. Next, we systematically analyze the performance deterioration of those fine-tuned models due to three distinct yet relevant types of input perturbations - oronyms, synonyms, and paraphrasing. Finally, we propose an efficient mitigation approach, Prompt Perturbation Consistency Learning (PPCL), which works by regularizing the divergence between losses from clean and perturbed samples. Our experiments demonstrate that PPCL can recover on average 59% and 69% of the performance drop for IC and SF tasks, respectively. Furthermore, PPCL beats the data augmentation approach while using ten times fewer augmented data samples.