| title: Capsule Pocket | |
| emoji: 💊 | |
| colorFrom: purple | |
| colorTo: yellow | |
| sdk: gradio | |
| sdk_version: "6.5.1" | |
| app_file: app.py | |
| pinned: false | |
| # Capsule Pocket | |
| Capsule Pocket trains seven primary capsules and ten eight-dimensional digit | |
| capsules with three rounds of routing by agreement. An ordinary MLP with exactly | |
| the same 4,060 trainable parameters is the control. | |
| The benchmark separates clean accuracy from one-pixel translation and center | |
| occlusion robustness. The Space exposes the ten digit-capsule vector lengths for | |
| each transformed input. | |
| ## Verified local result | |
| At exactly 4,060 parameters, capsules reached 97.04% clean accuracy versus 97.41% | |
| for the MLP. They improved one-pixel translation accuracy from 44.72% to 46.20% | |
| and center-occlusion accuracy from 79.63% to 82.22%. | |
| ```bash | |
| uv run python projects/capsule-pocket/train.py | |
| uv run pytest tests/test_capsule_pocket.py | |
| ``` | |