Frontier vision models from United States of America
19
Models
14 organisations
33%
Open weights
1 open, 2 closed
19
Frontier
By Epoch's flag
829M
Largest traceable
ResNeXt-101 32x48d
1960–2019
Span
10 with a traceable size
Organisation
Records
19| Model | Organisation | Released | Parameters | Weights |
|---|---|---|---|---|
| Noisy Student (L2) | Carnegie Mellon University (CMU) | 2019-11 | 480M | Closed |
| ResNeXt-101 32x48d | 2018-05 | 829M | Open | |
| JFT | Google Research | 2017-07 | 45M | Not recorded |
| NASv3 (CIFAR-10) | Google Brain | 2016-11 | 37M * | Not recorded |
| Xception | 2016-10 | 23M | Not recorded | |
| Inception v3 | 2015-12 | 24M * | Not recorded | |
| MSRA (C, PReLU) | Microsoft Research | 2015-02 | 87M | Not recorded |
| SPPNet | Microsoft | 2014-06 | — | Not recorded |
| Visualizing CNNs | New York University (NYU) | 2013-11 | — | Not recorded |
| Unsupervised High-level Feature Learner | 2012-07 | 1B * | Not recorded | |
| SVM-CNN | New York University (NYU) | 2006-06 | 90,857 | Closed |
| Decision tree (classification) | Mitsubishi Electric Research Labs | 2001-12 | 12,000 * | Not recorded |
| LeNet-5 | AT&T | 1998-11 | 60,000 | Not recorded |
| Siamese-TDNN | Bell Laboratories | 1993-08 | 744 * | Not recorded |
| Zip CNN | AT&T | 1989-12 | 9,760 | Not recorded |
| Handwritten digit recognition network | AT&T | 1989-11 | 2,578 | Not recorded |
| Print Recognition Logic | IBM | 1963-01 | — | Not recorded |
| ADALINE | Stanford University | 1960-06 | 17 | Not recorded |
| Perceptron (1960) | Cornell Aeronautical Laboratory | 1960-03 | 1,000 * | Not recorded |
An asterisk marks a parameter figure the source does not consider traceable to the people who built the model. That judgement is the source’s and it has false negatives: Kimi K3 is marked speculative at 2.8T while Moonshot’s own model card states the figure. Why that distinction carries the whole exhibit.
Counts are of models Epoch AI records as notable, not of every model released. The most recent month is always incomplete, because a model enters the dataset when it is reviewed rather than when it ships.
Epoch AI, 'Data on AI Models'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/data/ai-models-documentation' Used under CC BY 4.0.