Benchmarking Zero-shot Text Classification: Datasets, Evaluation and Entailment Approach
Open accessThe DOI matches its Crossref record; 32 of the 32 references it cites were found in a registry or an index.
Abstract
Wenpeng Yin, Jamaal Hay, Dan Roth. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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 is evidenced by LLMs impressive zero-shot and few-shot learning ability [18,132], as well as the fact that in-context learning does not require weight update, which mechanically prevents unfavorable fine-tuning-related behavior like catastrophic forgetting.”
From Harnessing the power of llms in practice: a survey on chatGPT and beyond (2024) · cites this paper as [132]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“LLMs have been shown to outperform previous zero-shot methods [132].”
From Harnessing the power of llms in practice: a survey on chatGPT and beyond (2024) · cites this paper as [132]
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
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173 · cited 6×
“Regarding this label-fully-unseen evaluation, apart from our entailment models and three unsupervised baselines “Majority”, “Word2Vec” and “ESA”, we also report the following baseline: Wikipedia-based : We train a binary clas-sifier based on BERT on a dataset collected from Wikipedia.”
MethodInfluentialBERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325 · cited 6×
“Regarding this label-fully-unseen evaluation, apart from our entailment models and three unsupervised baselines “Majority”, “Word2Vec” and “ESA”, we also report the following baseline: Wikipedia-based : We train a binary clas-sifier based on BERT on a dataset collected from Wikipedia.”
MethodInfluentialDistributed Representations of Words and Phrases and their Compositionality
Mikolov, T., Sutskever, I., Chen, K., Corrado, G., & Dean, J. (2013) · arXiv · doi:10.48550/arxiv.1310.4546 · Cited by 24,533 · cited 2×
“• Word2Vec 5 (Mikolov et al., 2013): Both the representations of the text and the labels are the addition of word embeddings element-wisely.”
BackgroundGLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding
Wang, A., Singh, A., Michael, J., Hill, F., Levy, O., & Bowman, S. (2018) · Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP · doi:10.18653/v1/w18-5446 · Cited by 4,222 · cited 11×
“For label-partially-unseen setup in which we intentionally provide annotated data, we first pretrain BERT on the MNLI/FEVER/RTE, then fine-tune on the provided training data.”
BackgroundMethodInfluentialA Broad-Coverage Challenge Corpus for Sentence Understanding through Inference
Williams, A., Nangia, N., & Bowman, S. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers) · doi:10.18653/v1/n18-1101 · Cited by 3,342 · cited 2×
“, 2019), and train it on three mainstream entailment datasets: MNLI (Williams et al., 2018), GLUE RTE (Dagan et al.”
BackgroundMethodCharacter-level Convolutional Networks for Text Classification
Zhang, X., Zhao, J., & LeCun, Y. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1509.01626 · Cited by 3,205 · cited 1×
“We use the large-scale Yahoo dataset released by Zhang et al. (2015).”
MethodComputing semantic relatedness using Wikipedia-based explicit semantic analysis
Gabrilovich, E., & Markovitch, S. (2007) · Cited by 1,981
The PASCAL Recognising Textual Entailment Challenge
Dagan, I., Glickman, O., & Magnini, B. (2006) · Lecture Notes in Computer Science · doi:10.1007/11736790_9 · Cited by 1,628
(2018) · International Conference on Learning Representations · Cited by 1,621
FEVER: a Large-scale Dataset for Fact Extraction and VERification
Thorne, J., Vlachos, A., Christodoulopoulos, C., & Mittal, A. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers) · doi:10.18653/v1/n18-1074 · Cited by 1,343 · cited 1×
“In this work, we make use of the widely-recognized state of the art entailment technique – BERT (Devlin et al., 2019), and train it on three mainstream entailment datasets: MNLI (Williams et al., 2018), GLUE RTE (Dagan et al., 2005; Wang et al., 2019) and FEVER 3 (Thorne et al., 2018), respectively.”
Method
Research works printed but not linked: 2
- [8]8. Chenliang Li, Jian Xing, Aixin Sun, and Zongyang Ma. 2016a. Effective document labeling with very few seed words: A topic model approach. In CIKM, pages 85-94.
- [11]11. Tomas Mikolov, Ilya Sutskever, Kai Chen, Gregory S. Corrado, and Jeffrey Dean. 2013. Distributed repre- sentations of words and phrases and their composi- tionality. In NIPS, pages 3111-3119.
Cited by
Items 501–600 of 867
EXnet: Efficient In-context Learning for Data-less Text classificationPreprint
Shome, D., & Yadav, K. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2305.14622 · Cited by 3
“Yin et al.(Yin et al., 2019) made the first attempt at unifying datasets and evaluating the zeroshot performance of language models over them which were only fine-tuned on natural language inference (NLI) tasks.”
BackgroundAnalysis of the Fed's communication by using textual entailment model of Zero-Shot classificationPreprint
Nakayama, Y., & Sawaki, T. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2306.04277 · Cited by 3
From a Large Language Model to Three-Dimensional SentimentPreprint
Clarke, P., Leininger, C., Principato, C., Staples, P. C., Goodwin, G. M., Ryslik, G., & Dougherty, R. F. (2023) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/kaeqy · Cited by 3
Jiang, Z., & Kim, D. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4782878 · Cited by 3
Healing Powers of BERT: How Task-Specific Fine-Tuning Recovers Corrupted Language ModelsPreprint
Han, S., Zhang, Z., & Simion, A. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2406.14459 · Cited by 3
“BERT’s rich pre-trained knowledge provides a strong foundation that can accurately classify text based on nuanced linguistic features when fine-tuned on labeled datasets (Yin et al., 2019).”
BackgroundText Psychometrics: Assessing Psychological Constructs in Text Using Natural Language ProcessingPreprint
Low, D. M., Mair, P., Nock, M. K., & Ghosh, S. (2025) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/9rdux_v2 · Cited by 3
Low, D. M. (2024) · Digital Access to Scholarship at Harvard (DASH) (Harvard University) · Cited by 3
Farzulla, M. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2601.20336 · Cited by 3
Grand Theft Identity: The Privacy Costs of DigitalizationPreprint
Phua, J. W. K., Wei, C., & Gloria Yang, Y. U. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5137239 · Cited by 3
VScript: Controllable Script Generation with Visual Presentation
Ji, Z., Xu, Y., Cheng, I.-T., Cahyawijaya, S., Frieske, R., Ishii, E., Zeng, M., Madotto, A., & Fung, P. (2022) · arXiv · doi:10.48550/arxiv.2203.00314 · Cited by 2
“In addition, we classify the video captions by zero-shot text classifier (Yin et al., 2019) to split this database based on genres.”
MethodThe Automatic Detection of Abusive Language in Dota 2 Chat Messages
Toit, J. L. D., & Kotzé, E. (2024) · International Conference on Artificial Intelligence, Computer, Data Sciences and Applications (ACDSA) · doi:10.1109/acdsa59508.2024.10467500 · Cited by 2
“This involved leveraging the capabilities of the "bart-large-mnli" pretrained language model [20] to assign labels to the unannotated dataset.”
MethodOne size fits all: Enhanced zero-shot text classification for patient listening on social media
Matoshi, V., Vuono, M. C. D., Gaspari, R., Kröll, M., Jantscher, M., Nicolardi, S. L., Mazzola, G., Rauch, M., Sabol, V., Salhofer, E., et al. (2025) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2024.1397470 · Cited by 2
Cheng, H., Shi, W., Sheng, B., Lee, A., Car, J., Chaudhary, V., Atanasov, A. G., Liu, N., Qiu, Y., Chen, Q., et al. (2025) · MedComm – Future Medicine · doi:10.1002/mef2.70019 · Cited by 2
The economic impact of uncertainty about U.S. regulations of the energy sector
Ma, X., & Xie, Z. (2025) · Economic Inquiry · doi:10.1111/ecin.13295 · Cited by 2
A Tool for Safe and Accurate IoT Automation Rule Generation Using Large Language Models
Islam, M. S., & Kantarcıoğlu, M. (2025) · IEEE Security and Privacy Workshops (SPW) · doi:10.1109/spw67851.2025.00025 · Cited by 2
Edwards, A., Pardiñas, A. F., Kirov, G., Rees, E., & Camacho-Collados, J. (2025) · JMIR AI · doi:10.2196/72256 · Cited by 2
“…are masked language models such as BERT (Bidirectional Encoder Representations From Transformers) and RoBERTa (Robustly Optimized BERT Approach) [7-12], which are pretrained on general-domain text and then fine-tuned— that is, further trained on labeled data for a specific down-stream task such…”
Latent space refinement for unsupervised cyber threat text classification
Wang, Y., Nayak, R., Bashar, M. A., & Chandramohan, M. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.114399 · Cited by 2
Using machine learning to automate the collection, transcription, and analysis of verbal-report data
Mechera‐Ostrovsky, T., Ungermann, P., & Donkin, C. (2025) · Behavior Research Methods · doi:10.3758/s13428-025-02800-5 · Cited by 2
“The text labeling employs a “zero shot” learning pipeline to assign labels to participant-recorded sentences (Yin et al., 2019; Sanh et al., 2022).”
Barone, S., Oggero, N., Damilano, M., & Battisti, E. (2025) · Technovation · doi:10.1016/j.technovation.2025.103420 · Cited by 2
Research on Optimizing Lightweight Small Models Based on Generating Training Data with ChatGPT
Ding, R., Zhu, E. Y., Zhao, C., Yang, H., Li, J., & Wu, Y. (2024) · Journal of Industrial Engineering and Applied Science · doi:10.5281/zenodo.10841042 · Cited by 2
Santos, E. C. S., & Han, H. D. (2026) · Futures & Foresight Science · doi:10.1002/ffo2.70026 · Cited by 2
Chatterjee, N., Koloski, B., Doucet, A., Pollak, S., & Purver, M. (2026) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2026.1844418 · Cited by 2
EDEntail: An Entailment-based Few-shot Text Classification with Extensional Definition
Zhu, Z., Qian, J., Feng, Z., Zhou, H., & Mao, K. (2024) · Findings of the Association for Computational Linguistics: NAACL · doi:10.18653/v1/2024.findings-naacl.71 · Cited by 2
“The intensional definition of a class label by WordNet, which specifies the necessary and sufficient conditions of the class label (Cook, 2009), is also frequently utilized to provide label information (Yin et al., 2019).”
BackgroundPatil, S., & Ravindran, B. (2024) · International Conference on Mining Software Repositories (MSR) · doi:10.1145/3643991.3645065 · Cited by 2
“We use the Deep Learning based Zero Shot (DL-ZS) text classifier proposed in [8] which has been pre-trained using natural language inference (NLI).”
MethodCui, X., Hanley, T., Choudhury, M., & Mu, T. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10651245 · Cited by 2
“As an alternative to the data-driven approach, we define it as a zero-shot multi-task classification problem [10], which is more flexible and challenging since the labels are unseen during model development.”
BackgroundMethodInfluentialQuestion Answering Module Leveraging Heterogeneous Datasets
Govindan, A., Ranjan, G., & Verma, A. (2021) · International Journal on Natural Language Computing · doi:10.5121/ijnlc.2021.10601 · Cited by 2
“As described in [26], the classifier was fine-tuned for the terms entailment, non-entailment, and neutral .”
Liu, Y., & Xie, X. (2022) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn55064.2022.9892830 · Cited by 2
“Yin et al. (2019) [17] construct a hypothesis by switching each candidate label and put the text to be recognized as the premise, thus turning the task of recognizing labels to the Natural Language Inference (NLI) task [18] and obtaining an out-of-the-box zero-shot recognition model that works well…”
BackgroundMethodInfluentialDig Here! Extracting and Using Knowledge from Financial Audit Reports
Pawar, S. S., Apte, M. M., Pawde, A., Vaishampayan, S., Palshikar, G. K., & Shinde, A. (2023) · Proceedings of the ... International Florida Artificial Intelligence Research Society Conference · doi:10.32473/flairs.36.133265 · Cited by 2
“0SHOT-TC: Zero-shot text classification (0SHOT-TC) (Yin, Hay, and Roth 2019) using Natural Language Inference (NLI) assigns a label to a text fragment even without its context such as domain, topic etc.”
QA-Matcher: Unsupervised Entity Matching Using a Question Answering Model
Hayashi, S., Dong, Y., & Oyamada, M. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-33383-5_14 · Cited by 2
Cross-task Knowledge Transfer for Extremely Weakly Supervised Text Classification
Park, S., Kim, K., & Lee, J. (2023) · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · doi:10.18653/v1/2023.findings-acl.328 · Cited by 2
“Our research bridges weakly-supervised noisy-label training with recent developments in prompt-based low-shot text classification (Yin et al., 2019; Keskar et al., 2019).”
BackgroundMethodNie, E., Schmid, H., & Schuetze, H. (2023) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2023.findings-emnlp.1056 · Cited by 2
Sulaiman, H. S., Gaidan, H. A., & Zamil, R. H. (2025) · Advancements in Life Sciences · doi:10.62940/als.v12i1.1944 · Cited by 2
Agrawal, A. (2025) · International Conference on Data Science and Machine Learning Applications (CDMA) · doi:10.1109/cdma61895.2025.00018 · Cited by 2
“While Scope 1 emissions dropped from 11.37 million metric tons CO 2 in 2018 to 9.19 million in 2020, Scope 2 emissions increased during the same period. his highlights the need for Walmart to further strengthen its environmental initiatives to match its social and governance commitments [18].”
InfluentialLi, W., Lalor, J. P., Chen, Y., & Kanuri, V. K. (2025) · ACM Transactions on Management Information Systems · doi:10.1145/3757747 · Cited by 2
“In the experiment, we utilize several unsupervised sentiment analysis tools (VADER (Hutto and Gilbert, 2014), TextBlob (Loria, 2018), and Zero-shot text classification (Yin et al., 2019)) to directly generate sentiment labels, which will replace the star rating labels given by users for training.”
THAURUS: An Innovative Multimodal Chatbot Based on the Next Generation of Conversational AI
Estecha-Garitagoitia, M., Rodríguez-Cantelar, M., Garrachón Ruiz, A., Fernández García, C. G., Esteban Romero, S., Conforto López, C., Fernández, A. S., Salvador, L. F. F., & Enríquez, L. F. D. (2023) · Archivo Digital UPM (Universidad Politécnica de Madrid) · Cited by 2
Benchmarking Conversation Routing in Chatbot Systems Based on Large Language Models
Maksymenko, D., Kryvoshein, D., Turuta, O., Kazakov, D., & Turuta, O. (2024) · International Workshop of IT-professionals on Artificial Intelligence (ProfIT AI) · doi:10.55056/ceur-ws.org/vol-3777/paper6.pdf · Cited by 2
“Natural Language Inference (NLI) models can be used as zero-shot classifiers [9].”
Charalampopoulos, D. K., & Koutsomitropoulos, D. A. (2022) · Communications in computer and information science · doi:10.1007/978-3-030-98876-0_23 · Cited by 2
“● A novel recommendation and scoring mechanism for subject tags that is based on zero-shot machine learning [20] and is therefore light on resources.”
BackgroundMethodDufour, S., Kandi, M. M., Boutamine, K., Gosse, C., Billami, M. B., Bortolaso, C., & Miloudi, Y. (2022) · International Workshop on Semantic Evaluation (SemEval) · doi:10.18653/v1/2022.semeval-1.173 · Cited by 2
“We used the zero-shot classification models (Yin, Hay and Roth, 2019) based on Press topics that we have defined manually: politics , sport , health , economy and technology .”
MethodTaxonomy-Guided Classification
Shen, J., & Han, J. (2022) · Springer eBooks · doi:10.1007/978-3-031-11405-2_5 · Cited by 2
Transformer Models for Activity Mining in Knowledge-Intensive Processes
Khandaker, F., Senderovich, A., Yu, E. S. K., Carbajales, S., & Chan, A. (2023) · Lecture notes in business information processing · doi:10.1007/978-3-031-25383-6_2 · Cited by 2
Kartchner, D., Al-Hussaini, I., Turner, H., Deng, J., Lohiya, S., Bathala, P., & Mitchell, C. S. (2023) · International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) · doi:10.1145/3539618.3591897 · Cited by 2
“[56] first propose approaching zero-shot text classification as a textual entailment problem.”
BackgroundPre-trained Models for Representation Learning
Lin, Y., Ding, N., Liu, Z., & Sun, M. (2023) · Springer eBooks · doi:10.1007/978-981-99-1600-9_5 · Cited by 2
Kyritsis, K., Spatiotis, N., Perikos, I., & Paraskevas, M. (2023) · Artificial intelligence · doi:10.5772/intechopen.112627 · Cited by 2
Aguilar, J. A. C., & Luna, E. (2024) · Latin American Business Review · doi:10.1080/10978526.2024.2320655 · Cited by 2
“Yin et al. (2019) making use of this type of task uses a pre-trained multi-genre sequence pair classifier (MNLI) as a ready-to-use zero-shot text classifier.”
MethodBarrière, V., & Jacquet, G. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.04052 · Cited by 2
“This can be seen as related to some works done on zero-shot learning using Natural Language Inference (NLI) where the semantic content of the label is encoded inside the classifier (Yin et al., 2020; Halder et al., 2020).”
BackgroundMethodFlexible Job Classification with Zero-Shot LearningPreprint
Lake, T. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2209.12678 · Cited by 2
“In [28] zero-shot document classification was formulated as an entailment task.”
MethodSanta, G. M., & Orozco‐Arroyave, J. R. (2023) · Research Square · doi:10.21203/rs.3.rs-3069667/v1 · Cited by 2
The 7th Workshop on Online Abuse and Harms (WOAH)
(2023) · doi:10.18653/v1/2023.woah-1 · Cited by 2
Is it possible for ChatGPT to mimic human annotator?Preprint
Kaszyca, O., Kazienko, P., Kocoń, J., Cichecki, I., Kochanek, M., & Szydło, D. (2023) · doi:10.36227/techrxiv.170259175.55207879/v1 · Cited by 2
Analysis of Socially Unacceptable Discourse with Zero-shot LearningPreprint
Ghilene, R., Niaouri, D., Linardi, M., & Longhi, J. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2409.13735 · Cited by 2
“In this regard, we showcase an illustrative example, drawing inspiration from prior research (Yin et al., 2019) which we have tailored to SUD analysis, as depicted in Figure 1.”
MethodGao, L. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2408.15650 · Cited by 2
Learning with Less: Knowledge Distillation from Large Language Models via Unlabeled DataPreprint
Li, J., Nag, S., Liu, H., Tang, X., Sarwar, S. M., Cui, L., Gu, H., Wang, S., He, Q. F., & Tang, J. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2411.08028 · Cited by 2
“A common approach is similarity-based methods (Abdullahi et al., 2024; Schopf et al., 2022; Yin et al., 2019), which generate embeddings for both input texts and labels, and then match texts to labels based on similarity.”
Text Psychometrics: Assessing Psychological Constructs in Text Using Natural Language ProcessingPreprint
Low, D. M., Mair, P., Nock, M. K., & Ghosh, S. (2025) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/9rdux_v4 · Cited by 2
A personality aware recommendation system
Elourajini, F. (2022) · Papyrus : Institutional Repository (Université de Montréal) · doi:10.71781/9799 · Cited by 2
Chang, H. H.-C., & Weener, T. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6162126 · Cited by 2
Think Twice Before You Write -- an Entropy-based Decoding Strategy to Enhance LLM ReasoningPreprint
He, J., Liu, M., Olaleye, O., Agarwal, A., Avendi, M., Abbasi, Y., Rowe, M., Patel, H. L., Li, P., Sheng, T., et al. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2604.00018 · Cited by 2
“An incorrect reasoning step may receive high probability due to patterns learned during pretraining, causing the beam to converge prematurely on flawed continuations while pruning the correct but lower probability trajectories (Yin et al., 2019).”
Self-citationRepresentation, information extraction, and summarization for automatic multimedia understanding
Harrando, I. (2022) · theses.fr (ABES) · doi:10.70675/702b8ba3z3acaz4ce3za9b9zf2cb1609e328 · Cited by 2
Behavioral Transfer in AI Agents: Evidence and Privacy ImplicationsPreprint
Luo, S., Zhang, Z., Dai, H., & Zhang, D. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2604.19925 · Cited by 2
Neural Router: Semantic Content Matching for Agentic AIPreprint
Lovén, L., Kumar, A., Engelhardt, A., Saleh, A., Morabito, R., Liu, X., Motlagh, N. H., & Tarkoma, S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2605.25701 · Cited by 2
“…( cross-encoder/ms-marco-MiniLM-L-6-v2 ) as a pairwise neural scoring model; zero-shot classification with DistilBART-MNLI-12-1 [Lewis et al. 2020; Yin et al. 2019] ( valhalla/distilbart-mnli-12-1 , distilled from facebook/bart-large-mnli ; events classified against subscription descriptions as…”
Learning to Classify Events from Human Needs Category Descriptions
Ding, H., & Feng, Z. (2020) · Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP) · doi:10.18653/v1/2020.findings-emnlp.421 · Cited by 1
“Entail : We also experimented with three pre-trained entailment models that are trained on: MNLI (Williams et al., 2018), GLUE RTE (Wang et al., 2018), FEVER (Thorne et al., 2018), and ˜tianyu/affEvent/affEventKB/ their ensemble model proposed in (Yin et al., 2019).”
BackgroundMethodInfluentialInjecting Commonsense Knowledge into Prompt Learning for Zero-Shot Text Classification
Qian, J., Chen, Q., Yue, Y., Atkinson, K. M., & Li, G. (2023) · International Conference on Machine Learning and Computing (ICMLC) · doi:10.1145/3587716.3587787 · Cited by 1
Adaptive Axes: A Pipeline for In-domain Social Stereotype Analysis
Zeng, Q., Jin, M., & Voigt, R. (2024) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2024.emnlp-main.872 · Cited by 1
Evaluating and Extending Techniques for Fine-Grained Text-Topic Prediction for Digital Forensic Data
Farabi, K. M. A., Agrawal, G., Dorai, G., Bardhan, R., Maleki, H., & Kadri, T. E. (2025) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-031-89360-5_16 · Cited by 1
Fat’hAlalim, A., Liu, Y., Xie, Q., Alsayed, A., & Eldow, M. (2025) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2025.0160579 · Cited by 1
Prompt-based Personality Profiling: Reinforcement Learning for Relevance Filtering
Hofmann, J., Sindermann, C., & Klinger, R. (2025) · Workshop for Research on Agent Language Models (REALM) · doi:10.18653/v1/2025.realm-1.1 · Cited by 1
Li, S., Mi, L., Castillo-Navarro, J., & Tuia, D. (2025) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2025.3620687 · Cited by 1
“The model is fine-tuned for natural language inference and used as a ready-made zero-shot sequence classifier [31].”
Sultana, J., & Rahman, R. M. (2025) · SN Computer Science · doi:10.1007/s42979-025-04445-9 · Cited by 1
“This approach involves the following steps: ● Model Utilization: The facebook/bart-large-mnli model [54], which is a pre-trained model known for its robust language understanding, was employed to classify the generated texts.”
Automated Scoring of Creative Achievement
Meinzer, N., Saretzki, J., Ceh, S. M., & Benedek, M. (2026) · The Journal of Creative Behavior · doi:10.1002/jocb.70104 · Cited by 1
Leveraging large language models for document classification in the banking sector
Nogueira, R., Mentzingen, H., & Garcia, N. C. (2026) · International Journal of Data Science and Analytics · doi:10.1007/s41060-026-01077-x · Cited by 1
Halimeh, H., Freese, F. K., & M�ller, O. (2025) · Proceedings of the International Conference on Information Systems Development · doi:10.62036/isd.2025.68 · Cited by 1
Sentiment analysis for software engineering: How far can zero-shot learning (ZSL) go?
Alfayez, R., & Binkhonain, M. (2025) · Information and Software Technology · doi:10.1016/j.infsof.2025.107971 · Cited by 1
“Moreover, our online appendix includes the full set of utilized labels 17 .”
Harnessing Language Models to Analyze Android App Permission Fidelity
Tamrakar, Y., Banerjee, R., Myers, E., Carli, L. D., & Ray, I. (2025) · Conference on Privacy, Security and Trust · doi:10.1109/pst65910.2025.11268849 · Cited by 1
“1.8T parameters, and its lightweight variant GPT-4o-mini, with 8B parameters— and compare them against established encoder-based base-lines: BERT-base (110M parameters) and BART-Large-MNLI (406M parameters) [30]–[32].”
153Detecting and Mapping Hate in Religious Contexts
Schneider, G., Goldzycher, J., & Volk, M. (2025) · De Gruyter eBooks · doi:10.1515/9783111721729-009 · Cited by 1
Automated Criterion-Based Thai Text Evaluation with Rationale and Improvement Guidance Using LLMs
Luantangsrisuk, V., Saengthongpattana, K., Chai, W. N., Boonsri, N., Choorat, T., Chunu, S., & Ruangrajitpakorn, T. (2025) · International Joint Symposium on Artificial Intelligence and Natural Language Processing (iSAI-NLP) · doi:10.1109/isai-nlp66160.2025.11320478 · Cited by 1
Feedback Intent Analysis using Machine Learning: A Case Study on Value Added Courses
Singh, S., Satsangee, D., & Singh, L. (2025) · International Conference on Information Systems and Computer Networks (ISCON) · doi:10.1109/iscon65210.2025.11340822 · Cited by 1
“It is a 406M parameter variant of BART with additional fine-tuning to perform zero-shot learning [22].”
Stars Are All You Need: A Distantly Supervised Pyramid Network for Unified Sentiment Analysis
Li, W., Chen, Y., Zheng, S., Wang, L., & Lalor, J. P. (2024) · Workshop on Noisy User-generated Text (W-NUT) · doi:10.18653/v1/2024.wnut-1.10 · Cited by 1
“In the experiment, we utilize several unsupervised sentiment analysis tools (VADER (Hutto and Gilbert, 2014), TextBlob (Loria, 2018), and Zero-shot text classification (Yin et al., 2019)) to directly generate sentiment labels, which will replace the star rating labels given by users for training.”
Enhancing Resume Screening Through Multi-stage LLM Classification and Hybrid Summarization
Jahan, N., Uddin, M. B., Arefin, M. S., Mehedi, C. M., Tasnim, F., & Delowar, K. E. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-15346-3_52 · Cited by 1
Computational Analysis of Contested Monuments and Collective Memory in a Multiethnic City
Nicula, A., Creţan, R., Drăgan, A., & Oancea, B. (2026) · Geographical Journal · doi:10.1111/geoj.70073 · Cited by 1
Exploring Sentiment Dynamics in the Bhagavad Gita Using Zero-Shot Learning and BART
Singh, V., Gaur, B., & Kumar, L. (2025) · OITS International Conference on Information Technology (OCIT) · doi:10.1109/ocit66168.2025.11400078 · Cited by 1
“Yin et al. [7] established a zero-shot classification method using BERT-based NLI models, showing its effectiveness for tasks like emotion recognition.”
Guo, X., Shan, G., Rivera, M., & Qiu, L. (2026) · Production and Operations Management · doi:10.1177/10591478261445688 · Cited by 1
“Surprise was estimated by assessing each comment’s semantic alignment with a predefined set of 20 organizational topic labels, where lower alignment indicated higher topical novelty (Yin et al., 2019).”
Iwaszenko, S., Czaja, S., & Kozłowski, A. (2026) · Applied Sciences · doi:10.3390/app16094562 · Cited by 1
Teaching AI to Decode Vaccine Hesitancy Narratives: A Few-Shot Learning and Topic Modeling Approach
Kabir, M. E., Tanim, S. H., Sellnow, D. D., Luteria, G. L. P., & Rennert, L. (2026) · Big Data and Cognitive Computing · doi:10.3390/bdcc10050159 · Cited by 1
“The zero-shot learning method, however, requires no task-specific training data and instead relies on natural language hypothesis templates to assign classifications [43].”
KvochurHegel at StanceNakba: Robust Stance Detection with Regularized Natural Language Inference
Le, M.-H. (2026) · Language Resources and Evaluation Conference (LREC) · doi:10.63317/3qqbomcjihz4 · Cited by 1
“To mitigate this behavior, we frame stance detection as an NLI task (Yin et al., 2019) using a DeBERTa-v3 Cross-Encoder.”
Anaraki, S. A. M., Croce, D., Costa, R., Tiburzi, L., Calabrese, A., & Basili, R. (2026) · Language Resources and Evaluation Conference (LREC) · doi:10.63317/4pebkuscwua5 · Cited by 1
“Recent efforts such as Climate-NLI (Yudanto et al., 2024) explore zero-and few-shot classification through natural language inference (Yin et al., 2019), yet performance is sensitive to semantic overlap and task ambiguity.”
Takahashi, H. (2026) · International Workshop on Semantic Evaluation (SemEval) · doi:10.18653/v1/2026.semeval-1.28 · Cited by 1
“Specifically, this experiment employed DeBERTa-base-long-nli (Sileo, 2024), RoBERTa-large-wanli (Liu et al., 2022), and BART-large-mnli (Lewis et al., 2020; Yin et al., 2019) models with prior fine-tuning on logical reasoning–related tasks.”
Why Low-Resource NLP Needs More Than Cross-Lingual Transfer: Lessons Learned from Luxembourgish
Philippy, F., Guo, S., Klein, J., & Bissyandé, T. F. (2026) · doi:10.18653/v1/2026.bigpicture-main.7 · Cited by 1
“Lesson 3: It Is Not Only How We Transfer, but What We Transfer A traditional and still widely adopted approach to zero-shot topic classification involves fine-tuning a model on Natural Language Inference (NLI) datasets (Yin et al., 2019).”
Kannojiya, V., & Dixit, V. (2026) · Procedia Computer Science · doi:10.1016/j.procs.2026.06.342 · Cited by 1
Liu, J., Luo, L., Pan, D., Kang, Y., Sun, Y., Lin, H., Yu, L., & Zhuang, H. (2026) · Health Information Science and Systems · doi:10.1007/s13755-026-00473-8 · Cited by 1
“The model then determines whether the premise entails the hypothesis [21].”
Unsupervised NLP for quantifying sentiment deviation and framing in global climate discourse
Aditya Vasudev, K., Saravanan, S., & Vinay, A. (2026) · Frontiers in Climate · doi:10.3389/fclim.2026.1827024 · Cited by 1
Meng, L. (2026) · eScholarship (California Digital Library) · Cited by 2
Jeong, M., & Yang, K. (2026) · Buildings · doi:10.3390/buildings16204131 · Cited by 1
Nasaq at AraGenre 2026: Definition-Guided Approaches for Hierarchical Arabic Genre Classification
Ahmad, S., Alharbi, A. I., Altamimi, S., Assiri, Y., & Alrajhi, M. (2026) · Arabic Natural Language Processing Workshop (WANLP) · doi:10.18653/2026.arabicnlp-sharedtasks.35 · Cited by 1
Narayan, R. P. (2026) · Natural Legal Language Processing Workshop (NLLP) · doi:10.18653/2026.nllp-1.8 · Cited by 1
Investigating Moral Foundations from Web Trending Topics
Tshimula, J. M., Chikhaoui, B., & Wang, S. (2022) · Lecture notes in networks and systems · doi:10.1007/978-3-031-14314-4_19 · Cited by 1
BOUN at SemEval-2021 Task 9: Text Augmentation Techniques for Fact Verification in Tabular Data
Köksal, A., Yüksel, Y., Yıldırım, B., & Özgür, A. (2021) · International Workshop on Semantic Evaluation (SemEval) · doi:10.18653/v1/2021.semeval-1.52 · Cited by 1
“…et al., 2005) is one of the core NLP problems for understanding the semantic relations between words and sentences, which is useful for other tasks including Question Answering (Abacha and Demner-Fushman, 2019), Text Summarization (Lloret et al., 2008), and Text Classification (Yin et al., 2019).”
CLZT: A Contrastive Learning Based Framework for Zero-Shot Text Classification
Li, K., Lin, M., Hu, S., & Li, R. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-00126-0_45 · Cited by 1
A Novel Hybrid Framework to Enhance Zero-shot Classification
Chen, Y., & Liu, Y. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-16270-1_17 · Cited by 1
Balaji, P., Merker, H., & Gupta, A. D. (2023) · JUCS - Journal of Universal Computer Science · doi:10.3897/jucs.84647 · Cited by 1
Improving Zero-Shot Inference with Unsupervised Key-Sentences Extraction
Pellicer, L. F. A. O., & Costa, A. H. R. (2023) · IEEE International Conference on Semantic Computing (ICSC) · doi:10.1109/icsc56153.2023.00024 · Cited by 1
“Some NLI models are applied to 0SHOT-TC, such as BART NLI [7] and DistilBERT.”
BackgroundMethodInfluential
Metadata
- Authors
- Yin, W., Hay, J., & Roth, D.
- Affiliations
- University of Pennsylvania (US)
- Venue
- Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
- Published
- 2019
- Pages
- 3912-3921
- OpenAlex
- W2970200208
- Cited by
- 537
- References
- 30
- Citations by year
- 2026: 66 · 2025: 86 · 2024: 75 · 2023: 141 · 2022: 94 · 2021: 64 · 2020: 11
- FWCI
- 16.17
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Conference on Empirical Methods in Natural Language Processing (EMNLP) (publishedVersion, cc-by)
- Funders
- U.S. Department of Defense
- Defense Advanced Research Projects Agency (HR0011-15-C-0113)
- Advanced Research Projects Agency
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Keywords
- zero-shot text classification, benchmarking, evaluation metrics, logical consequence, cross-domain classification
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.

