The DOI matches its Crossref record; 32 of the 32 references it cites were found in a registry or an index.
Demystifying Prompts in Language Models via Perplexity Estimation
Open accessAbstract
Language models can be prompted to perform a wide variety of tasks with zero-and few-shot incontext learning.However, performance varies significantly with the choice of prompt, and we do not yet understand why this happens.In this paper, we analyze the factors that contribute to this variance and establish a new empirical hypothesis: the performance of a prompt is predicted by the extent to which the model is familiar with the language it contains.Over a wide range of tasks, we show that the lower the perplexity of the prompt, the better it is able to perform the task, when considering reasonable prompts that are related to it.As part of our analysis, we also devise a method to automatically extend a small seed set of manually written prompts by paraphrasing with GPT3 and backtranslation.This larger set allows us to verify that perplexity is a strong predictor of the success of a prompt and we show that the lowest perplexity prompts are consistently effective.
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 source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This provides a clear signal for detection methods like Perplexity (Gonen et al., 2023) and Min-K% Prob (Shi et al., 2024), which are built upon this likelihood principle.”
From Detecting Data Contamination from Reinforcement Learning Post-training for Large Language Models (2025) · cites this paper as Gonen et al., 2023
Experimental — accuracy not yet measured
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
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
Character-level Convolutional Networks for Text ClassificationPreprint
Zhang, X., Zhao, J., & LeCun, Y. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1509.01626 · Cited by 4,481
DBpedia - A large-scale, multilingual knowledge base extracted from Wikipedia
Lehmann, J., Isele, R., Jakob, M., Jentzsch, A., Kontokostas, D., Mendes, P. N., Hellmann, S., Morsey, M., van Kleef, P., Auer, S., et al. (2015) · Semantic Web · doi:10.3233/sw-140134 · Cited by 3,737
Learning Word Vectors for Sentiment Analysis
Maas, A. L., Daly, R. E., Pham, P. T., Huang, D., Ng, A. Y., & Potts, C. (2011) · Cited by 3,710
The Power of Scale for Parameter-Efficient Prompt Tuning
Lester, B., Al-Rfou, R., & Constant, N. (2021) · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2021.emnlp-main.243 · Cited by 3,096
The Power of Scale for Parameter-Efficient Prompt Tuning
Lester, B., Al-Rfou, R., & Constant, N. (2021) · arXiv · doi:10.48550/arxiv.2104.08691 · Cited by 3,092
Prefix-Tuning: Optimizing Continuous Prompts for Generation
Li, X. L., & Liang, P. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.353 · Cited by 2,473
No Language Left Behind: Scaling Human-Centered Machine TranslationPreprint
Costa-jussà, M., Cross, J., Çelebi, O., Elbayad, M., Heafield, K., Heffernan, K., Kalbassi, E., Lam, J., Licht, D., Maillard, J., et al. (2022) · arXiv.org · doi:10.48550/arxiv.2207.04672 · Cited by 2,176
5.7 The WORDNET lexical database of English
Ooi, V. B. Y. (2019) · Edinburgh University Press eBooks · doi:10.1515/9781474471459-050 · Cited by 2,035
Making Pre-trained Language Models Better Few-shot Learners
Gao, T., Fisch, A., & Chen, D. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.295 · Cited by 1,380
AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts
Shin, T., Razeghi, Y., Logan IV, R. L., Wallace, E., & Singh, S. (2020) · Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2020.emnlp-main.346 · Cited by 1,279
Neural Network Acceptability Judgments
Warstadt, A., Singh, A., & Bowman, S. R. (2019) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00290 · Cited by 1,215
TweetEval: Unified Benchmark and Comparative Evaluation for Tweet Classification
Barbieri, F., Camacho-Collados, J., Anke, L. E., & Neves, L. (2020) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2020.findings-emnlp.148 · Cited by 929
It’s Not Just Size That Matters: Small Language Models Are Also Few-Shot Learners
Schick, T., & Schütze, H. (2021) · Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2021.naacl-main.185 · Cited by 747
RLPrompt: Optimizing Discrete Text Prompts with Reinforcement Learning
Deng, M., Wang, J., Hsieh, C.-P., Wang, Y., Guo, H. W., Shu, T., Song, M., Xing, E. P., & Hu, Z. (2022) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2022.emnlp-main.222 · Cited by 406
Estimating the Carbon Footprint of BLOOM, a 176B Parameter Language Model
Luccioni, A. S., Viguier, S., & Ligozat, A.-L. (2022) · arXiv · doi:10.48550/arxiv.2211.02001 · Cited by 378
Learning How to Ask: Querying LMs with Mixtures of Soft Prompts
Qin, G., & Eisner, J. (2021) · Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2021.naacl-main.410 · Cited by 342
CARER: Contextualized Affect Representations for Emotion Recognition
Saravia, E., Liu, H.-C. T., Huang, Y.-H., Wu, J., & Chen, Y.-S. (2018) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/d18-1404 · Cited by 337
How many data points is a prompt worth?
Le Scao, T., & Rush, A. (2021) · Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2021.naacl-main.208 · Cited by 179
PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts
Bach, S., Sanh, V., Yong, Z. X., Webson, A., Raffel, C., Nayak, N. V., Sharma, A., Kim, T., Bari, M. S., Fevry, T., et al. (2022) · Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics: System Demonstrations · doi:10.18653/v1/2022.acl-demo.9 · Cited by 178
Large Language Models Struggle to Learn Long-Tail Knowledge
Kandpal, N., Deng, H., Roberts, A., Wallace, E., & Raffel, C. (2022) · arXiv · doi:10.48550/arxiv.2211.08411 · Cited by 143
Reframing Instructional Prompts to GPTk’s Language
Mishra, S. R., Khashabi, D., Baral, C. R., Choi, Y., & Hajishirzi, H. (2022) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2022.findings-acl.50 · Cited by 140
Surface Form Competition: Why the Highest Probability Answer Isn’t Always Right
Holtzman, A., West, P., Shwartz, V., Choi, Y., & Zettlemoyer, L. (2021) · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2021.emnlp-main.564 · Cited by 127
Self-citationOn the Relation between Sensitivity and Accuracy in In-Context Learning
Chen, Y., Zhao, C., Zhou, Y., McKeown, K. R., & He, H. (2023) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2023.findings-emnlp.12 · Cited by 62
Prompt Waywardness: The Curious Case of Discretized Interpretation of Continuous Prompts
Khashabi, D., Lyu, X., Min, S., Qin, L., Richardson, K., Welleck, S., Hajishirzi, H., Khot, T., Sabharwal, A., Singh, S., et al. (2022) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/2022.naacl-main.266 · Cited by 58
Impact of Pretraining Term Frequencies on Few-Shot Reasoning
Razeghi, Y., Logan, R. L., Gardner, M., & Singh, S. (2022) · arXiv · doi:10.48550/arxiv.2202.07206 · Cited by 53
NorthEuraLex: a wide-coverage lexical database of Northern Eurasia
Dellert, J., Daneyko, T., Münch, A., Ladygina, A., Buch, A., Clarius, N., Grigorjew, I., Balabel, M., Boga, H. I., Baysarova, Z., et al. (2019) · Language Resources and Evaluation · doi:10.1007/s10579-019-09480-6 · Cited by 49
Toward Human Readable Prompt Tuning: Kubrick’s The Shining is a good movie, and a good prompt too?
Shi, W., Han, X., Gonen, H., Holtzman, A., Tsvetkov, Y., & Zettlemoyer, L. (2023) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2023.findings-emnlp.733 · Cited by 44
Self-citationMulti-Source Social Feedback of Online News FeedsPreprint
Moniz, N., & Torgo, L. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1801.07055 · Cited by 39
Measuring Causal Effects of Data Statistics on Language Model's `Factual' PredictionsPreprint
Elazar, Y., Kassner, N., Ravfogel, S., Feder, A., Ravichander, A., Mosbach, M., Belinkov, Y., Schütze, H., & Goldberg, Y. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2207.14251 · Cited by 15
Han, X., & Tsvetkov, Y. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2205.12600 · Cited by 10
OPT: Open Pre-trained Transformer Language Models
Zhang, S., Roller, S., Goyal, N., Artetxe, M., Chen, M., Chen, S., Dewan, C., Diab, M., Li, X., Lin, X. V., et al. (2022) · arXiv · doi:10.48550/arxiv.2205.01068
Self-citation
Research works printed but not linked: 2
- George A Miller. 1995. Wordnet: a lexical database for english. Communications of the ACM, 38(11):39-41.
- John Wieting, Jonathan Mallinson, and Kevin Gimpel. 2017. Learning paraphrastic sentence embeddings from back-translated bitext. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 274-285.
Cited by
Sorted by citations
A Survey on In-context Learning
Dong, Q., Li, L., Dai, D., Zheng, C., Ma, J., Li, R., Xia, H., Xu, J., Wu, Z., Chang, B., et al. (2024) · Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2024.emnlp-main.64 · Cited by 649
Security and Privacy Challenges of Large Language Models: A Survey
Das, B. C., Amini, M. H., & Wu, Y. (2025) · ACM Computing Surveys · doi:10.1145/3712001 · Cited by 445
Gao, Y., Li, R., Croxford, E., Caskey, J., Patterson, B. W., Churpek, M. M., Miller, T. A., Dligach, D., & Afshar, M. (2024) · JMIR AI · doi:10.2196/58670 · Cited by 169
State of What Art? A Call for Multi-Prompt LLM Evaluation
Mizrahi, M., Kaplan, G., Malkin, D., Dror, R., Shahaf, D., & Stanovsky, G. (2024) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00681 · Cited by 113
Prompting meaning: a hermeneutic approach to optimising prompt engineering with ChatGPT
Henrickson, L., & Meroño-Peñuela, A. (2023) · AI & Society · doi:10.1007/s00146-023-01752-8 · Cited by 136
Metadata
- Authors
- Gonen, H., Iyer, S., Blevins, T., Smith, N., & Zettlemoyer, L.
- Affiliations
- University of Washington (US)
- Allen Institute for Artificial Intelligence (US)
- Venue
- Findings of the Association for Computational Linguistics: EMNLP 2023
- Published
- 2023
- Pages
- 10136-10148
- OpenAlex
- W4389520255
- Cited by
- 109
- References
- 32
- Citations by year
- 2026: 14 · 2025: 33 · 2024: 30 · 2023: 31
- FWCI
- 19.14
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) (publishedVersion, cc-by)
- 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 engineering, language models, in-context learning, zero-shot learning, few-shot learning, paraphrase, back-translation
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
