Next-token Prediction
Next-Token Prediction is a computational task in language modeling where the objective is to predict the most probable subsequent token in a sequence given its preceding context, serving as the foundational training paradigm for autoregressive language models.
68
Publications
7.3K
Citations
1.1K
Authors
83
Institutions
Publications per year
2017–2026
63
Variants
Next Token Prediction
68
arXiv (Cornell University) · 2023 · 2.4K citations · Full text
Federated Learning for Mobile Keyboard Prediction
Andrew Hard, K. Praveen Kumar Rao, Rajiv Mathews et al. · arXiv (Cornell University) · 2018 · 1.1K citations · Full text
Toolformer: Language Models Can Teach Themselves to Use Tools
Timo Schick, Jane Dwivedi-Yu, Roberto Dessì et al. · arXiv (Cornell University) · 2023 · 393 citations · Full text
ProphetNet: Predicting Future N-gram for Sequence-to-SequencePre-training
Weizhen Qi, Yu Yan, Yeyun Gong et al. · 2020 · 340 citations · Full text
1–5 of 68
1.1K
2
h-index
2
h-index
2
h-index
2
h-index
2
h-index
1–5 of 1.1K
83
Princeton University
Princeton, United States
3 h-index
11 publications
482 citations
Stanford University
Stanford, United States
3 h-index
6 publications
145 citations
Columbia University
New York, United States
2 h-index
3 publications
220 citations
New York University
New York, United States
2 h-index
2 publications
65 citations

Peking University
Beijing, China
2 h-index
5 publications
681 citations
International Institute of Information Technology, Hyderabad
Hyderabad, India
2 h-index
6 publications
87 citations
University College London
London, United Kingdom
2 h-index
8 publications
205 citations

Zhejiang University
Hangzhou, China
2 h-index
13 publications
453 citations
1–8 of 83
Repository · 29 publications · 5.4K citations
20
h-index
Repository · 2 publications · 29 citations
2
h-index
1 publications · 114 citations
1
h-index
1 publications · 24 citations
1
h-index
Conference · 1 publications · 14 citations
1
h-index
1–5