| natural-language-understanding-on-glue-dev | Perceiver IO++ | #12 | GLUE (dev): 81.8 |
| optical-flow-estimation-on-sintel-final | Perceiver IO | #13 | Average End-Point Error: 2.42 |
| audio-tagging-on-audioset | Perceiver IO (mel-spectrogram + video) | #14 | mean average precision: 0.450 |
| optical-flow-estimation-on-sintel-clean | Perceiver IO | #22 | Average End-Point Error: 1.81 |
| audio-classification-on-audioset | Perceiver IO (mel-spectrogram + video) | #35 | Test mAP: 0.449 |
| optical-flow-estimation-on-kitti-2015 | Perceiver IO | #38 | Average End-Point Error: 4.98 |
| image-classification-on-imagenet | Perceiver IO (config B, pretrained) | #333 | Top 1 Accuracy: 84.5% |