next-steps.md

For the GATE paper:

1. Image classification -> Done

2. Visual relational reasoning -> Done

3. Semantic segmentation

4. Few Shot Learning -> Working on it

5. Zero shot learning

6. Medical image classification

7. Medical semantic segmentation

8. Video classification

9. Text classification (models that support text modalities, CLIP, other multi-modal foundation models)

10. Audio classification -> as a modality shift (remove root, replace with new modality root embedding)

Premise is that we think, based on recent and past work, that we need a more