Fine-Tuning on Noisy Instructions: Effects on Generalization and Performance
Not on the map. Why?
Abstract (arXiv)
Instruction-tuning plays a vital role in enhancing the task-solving abilities of large language models (LLMs), improving their usability in generating helpful responses on various tasks. However, previous work has demonstrated that they are sensitive to minor variations in instruction phrasing. In this paper, we explore whether introducing perturbations in instruction-tuning data can enhance LLMs' resistance against noisy instructions. We focus on how instruction-tuning with perturbations, such as removing stop words or shuffling words, affects LLMs' performance on the original and perturbed versions of widely-used benchmarks (MMLU, BBH, GSM8K). We further assess learning dynamics and potential shifts in model behavior. Surprisingly, our results suggest that instruction-tuning on perturbed instructions can, in some cases, improve downstream performance. These findings highlight the importance of including perturbed instructions in instruction-tuning, which can make LLMs more resilient to noisy user inputs.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 32 of its 49 references that are in the map. First posted on arXiv 3 Oct 2025.
Builds on
32 of 49 references.
References · 32
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
- 2022Training language models to follow instructions with human feedbackcited by 15,712
- 2024The Llama 3 Herd of Modelscited by 14,192
- 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
- 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
- 2021Training Verifiers to Solve Math Word Problemscited by 9,743
- 2020Measuring Massive Multitask Language Understandingcited by 7,768
- 2022Large Language Models are Zero-Shot Reasonerscited by 4,323
- 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
- 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,355
- 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,919
- 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,767
20 more
- 2022Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Taskscited by 858
- 2021Cross-Task Generalization via Natural Language Crowdsourcing Instructionscited by 696
- 2024OLMo: Accelerating the Science of Language Modelscited by 599
- 2022ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detectioncited by 419
- 2023How Far Can Camels Go? Exploring the State of Instruction Tuning on Open Resourcescited by 405
- 2023The Unlocking Spell on Base LLMs: Rethinking Alignment via In-Context Learningcited by 251
- 2024Unpacking DPO and PPO: Disentangling Best Practices for Learning from Preference Feedbackcited by 107
- 2018A Kernel Theory of Modern Data Augmentationcited by 106
- 2018Data augmentation instead of explicit regularizationcited by 62
- 2024Learning or Self-aligning? Rethinking Instruction Fine-tuningcited by 43
- 2019Improving Robustness of Machine Translation with Synthetic Noisecited by 36
- 2024Improving the Robustness of Large Language Models via Consistency Alignmentcited by 36
- 2022Robustness of Learning from Task Instructionscited by 34
- 2023MUFFIN: Curating Multi-Faceted Instructions for Improving Instruction-Followingcited by 30
- 2023Instructive Decoding: Instruction-Tuned Large Language Models are Self-Refiner from Noisy Instructionscited by 24
- 2024RobustFT: Robust Supervised Fine-tuning for Large Language Models under Noisy Responsecited by 24
- 2024Resilience of Large Language Models for Noisy Instructionscited by 20
- 2024Contrastive Instruction Tuningcited by 17
- 2025Enhancing LLM Robustness to Perturbed Instructions: An Empirical Studycited by 17
- 2025ArithmAttack: Evaluating Robustness of LLMs to Noisy Context in Math Problem Solvingcited by 13
Built on by
Citing papers per year: 2026: 2.
2 papers.