{"url":"/method/mpnet","slug":"mpnet","name":"MPNet","full_name":"MPNet","full_name_withheld":false,"description_markdown":"**MPNet** is a pre-training method for language models that combines masked language modeling (MLM) and permuted language modeling (PLM) in one view. It takes the dependency among the predicted tokens into consideration through permuted language modeling and thus avoids the issue of [BERT](https://paperswithcode.com/method/bert). On the other hand, it takes position information of all tokens as input to make the model see the position information of all the tokens and thus alleviates the position discrepancy of [XLNet](https://paperswithcode.com/method/xlnet).\r\n\r\nThe training objective of MPNet is:\r\n\r\n$$ \\mathbb{E}\\_{z\\in{\\mathcal{Z}\\_{n}}} \\sum^{n}\\_{t=c+1}\\log{P}\\left(x\\_{z\\_{t}}\\mid{x\\_{z\\_{<t}}}, M\\_{z\\_{{>}{c}}}; \\theta\\right) $$\r\n\r\nAs can be seen, MPNet conditions on ${x\\_{z\\_{<t}}}$ (the tokens preceding the current predicted token $x\\_{z\\_{t}}$) rather than only the non-predicted tokens ${x\\_{z\\_{<=c}}}$ in MLM; comparing with PLM, MPNet takes more information (i.e., the mask symbol $[M]$ in position $z\\_{>c}$) as inputs. Although the objective seems simple, it is challenging to implement the model efficiently. For details, see the paper.","description_state":"present","introduced_year":null,"introduced_by":{"title":"MPNet: Masked and Permuted Pre-training for Language Understanding","paper":"/paper/mpnet-masked-and-permuted-pre-training-for","first_author":"Kaitao Song","n_authors":5,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/mpnet-masked-and-permuted-pre-training-for"},"source":{"url":"https://arxiv.org/abs/2004.09297v2","title":"MPNet: Masked and Permuted Pre-training for Language Understanding","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"Natural Language Processing","area_id":"natural-language-processing","collection":"Language Model Pre-Training","url":"/methods/category/language-model-pre-training","pwc_aliases":[]}],"n_papers_tagged":17,"archive_num_papers":17,"papers_newest_first":[{"paper":null,"title":"Computational Detection of Intertextual Parallels in Biblical Hebrew: A Benchmark Study Using Transformer-Based Language Models","date":"2025-06-30","arxiv_id":"2506.24117","n_code_links":0,"syntology":null},{"paper":null,"title":"Large Language Model Guided Progressive Feature Alignment for Multimodal UAV Object Detection","date":"2025-03-10","arxiv_id":"2503.06948","n_code_links":0,"syntology":null},{"paper":null,"title":"\"Actionable Help\" in Crises: A Novel Dataset and Resource-Efficient Models for Identifying Request and Offer Social Media Posts","date":"2025-02-24","arxiv_id":"2502.16839","n_code_links":0,"syntology":null},{"paper":null,"title":"Explicit Depth-Aware Blurry Video Frame Interpolation Guided by Differential Curves","date":"2025-01-01","arxiv_id":null,"n_code_links":0,"syntology":null},{"paper":null,"title":"CReMa: Crisis Response through Computational Identification and Matching of Cross-Lingual Requests and Offers Shared on Social Media","date":"2024-05-20","arxiv_id":"2405.11897","n_code_links":0,"syntology":null},{"paper":null,"title":"Harnessing PubMed User Query Logs for Post Hoc Explanations of Recommended Similar Articles","date":"2024-02-05","arxiv_id":"2402.03484","n_code_links":0,"syntology":null},{"paper":null,"title":"RECipe: Does a Multi-Modal Recipe Knowledge Graph Fit a Multi-Purpose Recommendation System?","date":"2023-08-08","arxiv_id":"2308.04579","n_code_links":0,"syntology":null},{"paper":"/paper/specious-sites-tracking-the-spread-and-sway","title":"Specious Sites: Tracking the Spread and Sway of Spurious News Stories at Scale","date":"2023-08-03","arxiv_id":"2308.02068","n_code_links":1,"syntology":null},{"paper":"/paper/identifying-misinformation-on-youtube-through","title":"Identifying Misinformation on YouTube through Transcript Contextual Analysis with Transformer Models","date":"2023-07-22","arxiv_id":"2307.12155","n_code_links":1,"syntology":null},{"paper":null,"title":"Utilizing ChatGPT Generated Data to Retrieve Depression Symptoms from Social Media","date":"2023-07-05","arxiv_id":"2307.02313","n_code_links":0,"syntology":null},{"paper":null,"title":"Vec2Vec: A Compact Neural Network Approach for Transforming Text Embeddings with High Fidelity","date":"2023-06-22","arxiv_id":"2306.12689","n_code_links":0,"syntology":null},{"paper":null,"title":"Partial Mobilization: Tracking Multilingual Information Flows Amongst Russian Media Outlets and Telegram","date":"2023-01-25","arxiv_id":"2301.10856","n_code_links":0,"syntology":null},{"paper":null,"title":"Using Large Pre-Trained Language Model to Assist FDA in Premarket Medical Device","date":"2022-11-03","arxiv_id":"2212.01217","n_code_links":0,"syntology":null},{"paper":null,"title":"Happenstance: Utilizing Semantic Search to Track Russian State Media Narratives about the Russo-Ukrainian War On Reddit","date":"2022-05-28","arxiv_id":"2205.14484","n_code_links":0,"syntology":null},{"paper":"/paper/youngsheldon-at-semeval-2021-task-5-fine","title":"YoungSheldon at SemEval-2021 Task 5: Fine-tuning Pre-trained Language Models for Toxic Spans Detection using Token classification Objective","date":"2021-08-01","arxiv_id":null,"n_code_links":1,"syntology":null},{"paper":"/paper/mpnet-variable-depth-unfolded-neural-network","title":"mpNet: variable depth unfolded neural network for massive MIMO channel estimation","date":"2020-08-07","arxiv_id":"2008.04088","n_code_links":1,"syntology":null},{"paper":"/paper/mpnet-masked-and-permuted-pre-training-for","title":"MPNet: Masked and Permuted Pre-training for Language Understanding","date":"2020-04-20","arxiv_id":"2004.09297","n_code_links":7,"syntology":{"ran":5,"of":8,"unverified":3,"pointer_only":6}}],"papers_shown":17,"tasks":[{"task":"/task/sentence","name":"Sentence","papers":7},{"task":"/task/language-modelling","name":"Language Modelling","papers":5},{"task":"/task/articles","name":"Articles","papers":4},{"task":"/task/language-modeling","name":"Language Modeling","papers":4},{"task":"/task/large-language-model","name":"Large Language Model","papers":3},{"task":"/task/misinformation","name":"Misinformation","papers":2},{"task":"/task/recommendation-systems","name":"Recommendation Systems","papers":2},{"task":"/task/word-embeddings","name":"Word Embeddings","papers":2},{"task":"/task/benchmarking","name":"Benchmarking","papers":1},{"task":"/task/classification-1","name":"Classification","papers":1},{"task":"/task/collaborative-filtering","name":"Collaborative Filtering","papers":1},{"task":"/task/fact-checking","name":"Fact Checking","papers":1},{"task":"/task/few-shot-learning","name":"Few-Shot Learning","papers":1},{"task":"/task/graph-embedding","name":"Graph Embedding","papers":1},{"task":"/task/knowledge-graph-embedding","name":"Knowledge Graph Embedding","papers":1},{"task":"/task/knowledge-graphs","name":"Knowledge Graphs","papers":1},{"task":"/task/marketing","name":"Marketing","papers":1},{"task":"/task/masked-language-modeling","name":"Masked Language Modeling","papers":1},{"task":"/task/multi-modal-knowledge-graph","name":"Multi-modal Knowledge Graph","papers":1},{"task":"/task/object","name":"Object","papers":1}],"tasks_shown":20,"n_tasks":38,"usage_by_year":[{"year":"2020","papers":2},{"year":"2021","papers":1},{"year":"2022","papers":2},{"year":"2023","papers":6},{"year":"2024","papers":2},{"year":"2025","papers":4}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/mpnet"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}