The DOI matches its Crossref record; 1 of the 1 references it cites were found in a registry or an index.
ProSA: Assessing and Understanding the Prompt Sensitivity of LLMs
Open accessAbstract
Large language models (LLMs) have demonstrated impressive capabilities across various tasks, but their performance is highly sensitive to the prompts utilized.This variability poses challenges for accurate assessment and user satisfaction.Current research frequently overlooks instance-level prompt variations and their implications on subjective evaluations.To address these shortcomings, we introduce ProSA, a framework designed to evaluate and comprehend prompt sensitivity in LLMs.ProSA incorporates a novel sensitivity metric, PromptSensiScore, and leverages decoding confidence to elucidate underlying mechanisms.Our extensive study, spanning multiple tasks, uncovers that prompt sensitivity fluctuates across datasets and models, with larger models exhibiting enhanced robustness.We observe that few-shot examples can alleviate this sensitivity issue, and subjective evaluations are also susceptible to prompt sensitivities, particularly in complex, reasoningoriented tasks.Furthermore, our findings indicate that higher model confidence correlates with increased prompt robustness.We believe this work will serve as a helpful tool in studying prompt sensitivity of LLMs.The project is released at: https://github.com/ open-compass/ProSA.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“LLM responses are known to be highly sensitive to even small variations in prompt wording Zhuo et al. (2024).”
From Challenging the Evaluator: LLM Sycophancy Under User Rebuttal (2025) · cites this paper as Zhuo et al. (2024)
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
The publisher hasn’t deposited this paper’s reference list.
Research works printed but not linked: 21
- Jacob Austin, Augustus Odena, Maxwell Nye, Maarten Bosma, Henryk Michalewski, David Dohan, Ellen Jiang, Carrie Cai, Michael Terry, Quoc Le, and Charles Sutton. 2021. Program synthesis with large language models. Preprint, arXiv:2108.07732. · arXiv
- Philippe Tillet, Felipe Petroski Such, Dave Cum- mings, Matthias Plappert, Fotios Chantzis, Eliza- beth Barnes, Ariel Herbert-Voss, William Hebgen Guss, Alex Nichol, Alex Paino, Nikolas Tezak, Jie Tang, Igor Babuschkin, Suchir Balaji, Shantanu Jain, William Saunders, Christopher Hesse, Andrew N. Carr, Jan Leike, Josh Achiam, Vedant Misra, Evan Morikawa, Alec Radford, Matthew Knight, Miles Brundage, Mira Murati, Katie Mayer, Peter Welinder, Bob McGrew, Dario Amodei, Sam McCandlish, Ilya Sutskever, and Wojciech Zaremba. 2021. Evaluating large language models trained on code.
- Peter Clark, Isaac Cowhey, Oren Etzioni, Tushar Khot, Ashish Sabharwal, Carissa Schoenick, and Oyvind Tafjord. 2018. Think you have solved question answering? try arc, the ai2 reasoning challenge. Preprint, arXiv:1803.05457. · arXiv
- Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. 2021. Training verifiers to solve math word prob- lems. arXiv preprint arXiv:2110.14168. · arXiv
- Yann Dubois, Balázs Galambosi, Percy Liang, and Tat- sunori B Hashimoto. 2024. Length-controlled al- pacaeval: A simple way to debias automatic evalua- tors. arXiv preprint arXiv:2404.04475. · arXiv
- Jiahui Geng, Fengyu Cai, Yuxia Wang, Heinz Koeppl, Preslav Nakov, and Iryna Gurevych. 2023. A sur- vey of language model confidence estimation and calibration. arXiv preprint arXiv:2311.08298. · arXiv
- Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. 2021. Measuring mathematical problem solving with the MATH dataset. In Thirty- fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2).
- Bonan Min, Hayley Ross, Elior Sulem, Amir Pouran Ben Veyseh, Thien Huu Nguyen, Oscar Sainz, Eneko Agirre, Ilana Heintz, and Dan Roth. 2023. Recent advances in natural language processing via large pre-trained language models: A survey. ACM Computing Surveys, 56(2):1-40.
- Moran Mizrahi, Guy Kaplan, Dan Malkin, Rotem Dror, Dafna Shahaf, and Gabriel Stanovsky. 2024. State of what art? a call for multi-prompt llm evaluation. CoRR, abs/2401.00595. · DOI
- Nasrin Mostafazadeh, Michael Roth, Annie Louis, Nathanael Chambers, and James Allen. 2017. Ls- dsem 2017 shared task: The story cloze test. In Proceedings of the 2nd Workshop on Linking Models of Lexical, Sentential and Discourse-level Semantics, pages 46-51.
- Pouya Pezeshkpour and Estevam Hruschka. 2024. Large language models sensitivity to the order of op- tions in multiple-choice questions. In Findings of the Association for Computational Linguistics: NAACL 2024, pages 2006-2017, Mexico City, Mexico. Asso- ciation for Computational Linguistics. · DOI · Paper page
- Nils Reimers and Iryna Gurevych. 2019. Sentence-bert: Sentence embeddings using siamese bert-networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing. Associa- tion for Computational Linguistics.
- Keisuke Sakaguchi, Ronan Le Bras, Chandra Bhagavat- ula, and Yejin Choi. 2021. Winogrande: An adver- sarial winograd schema challenge at scale. Commu- nications of the ACM, 64(9):99-106.
- Tal Schuster, Adam Fisch, Jai Gupta, Mostafa Dehghani, Dara Bahri, Vinh Tran, Yi Tay, and Donald Metzler. 2022. Confident adaptive language modeling. Ad- vances in Neural Information Processing Systems, 35:17456-17472.
- Melanie Sclar, Yejin Choi, Yulia Tsvetkov, and Alane Suhr. 2024. Quantifying language models' sensitiv- ity to spurious features in prompt design or: How i learned to start worrying about prompt formatting. In The Twelfth International Conference on Learning Representations.
- Alon Talmor, Jonathan Herzig, Nicholas Lourie, and Jonathan Berant. 2019. CommonsenseQA: A ques- tion answering challenge targeting commonsense knowledge. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Tech- nologies, Volume 1 (Long and Short Papers), pages 4149-4158, Minneapolis, Minnesota. Association for Computational Linguistics. · DOI
- Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q. Weinberger, and Yoav Artzi. 2020. Bertscore: Eval- uating text generation with bert. In International Conference on Learning Representations.
- Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, et al. 2023. A survey of large language models. arXiv preprint arXiv:2303.18223. · arXiv
- Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Tianle Li, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zhuohan Li, Zi Lin, Eric Xing, et al. 2023. Lmsys- chat-1m: A large-scale real-world llm conversation dataset. arXiv preprint arXiv:2309.11998. · arXiv
- Kun Zhou, Yutao Zhu, Zhipeng Chen, Wentong Chen, Wayne Xin Zhao, Xu Chen, Yankai Lin, Ji-Rong Wen, and Jiawei Han. 2023. Don't make your llm an evaluation benchmark cheater. arXiv preprint arXiv:2311.01964. · arXiv
- Kaijie Zhu, Jindong Wang, Jiaheng Zhou, Zichen Wang, Hao Chen, Yidong Wang, Linyi Yang, Wei Ye, Neil Zhenqiang Gong, Yue Zhang, et al. 2023. Promptbench: Towards evaluating the robustness of large language models on adversarial prompts. arXiv preprint arXiv:2306.04528. · arXiv
Cited by
Sorted by citations
Benchmarking Prompt Sensitivity in Large Language Models
Razavi, A., Soltangheis, M., Arabzadeh, N., Salamat, S., Zihayat, M., & Bagheri, E. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-88714-7_29 · Cited by 78
Brittlebench: Quantifying LLM robustness via prompt sensitivity
Romanou, A., Ibrahim, M., Ross, C., Oktar, K., Shaib, C., Bell, S. J., Ovalle, A., Bosselut, A., Dodge, J., Sinha, K., et al. (2026) · arXiv · doi:10.48550/arxiv.2603.13285 · Cited by 24
Inherent and emergent liability issues in LLM-based agentic systems: a principal-agent perspective
Gabison, G. A., & Xian, R. P. (2025) · Proceedings of the 1st Workshop for Research on Agent Language Models (REALM 2025) · doi:10.18653/v1/2025.realm-1.9 · Cited by 19
Randomness, Not Representation: The Unreliability of Evaluating Cultural Alignment in LLMs
Khan, A., Casper, S., & Hadfield-Menell, D. (2025) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3715275.3732147 · Cited by 29
de Oliveira, R., Garber, M., Gwinnutt, J. M., Rashidi, E., Hwang, J.-H. (., Gilmour, W., Nanavati, J., Abidine, K. Z. E., & Mack, C. D. (2025) · JAMIA Open · doi:10.1093/jamiaopen/ooaf058 · Cited by 22
Metadata
- Authors
- Zhuo, J., Zhang, S., Fang, X., Duan, H., Lin, D., & Chen, K.
- Affiliations
- Jilin University (CN)
- Shanghai Artificial Intelligence Laboratory (CN)
- Chinese University of Hong Kong (HK)
- Venue
- Findings of the Association for Computational Linguistics: EMNLP 2024
- Published
- 2024
- Pages
- 1950-1976
- OpenAlex
- W4404782660
- Cited by
- 78
- References
- 0
- Citations by year
- 2026: 54 · 2025: 24
- FWCI
- 22.52
- Type
- conference paper
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by)
- arXiv (Cornell University) (cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- prompt sensitivity, large language models, prompt robustness, subjective evaluation, few-shot prompting, model size, reasoning tasks
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
