Requirements

Students should work in teams of no more than five members. Each project should involve at least two modalities. VLA models are considered part of multimodal large models.

Project Options

1. Complete a project related to multimodal large models.
2. Write an innovation proposal on multimodal large models.
3. Write or reproduce a paper targeting top AI conferences, such as NeurIPS, EMNLP, or CoRL.
4. Participate in a global competition related to multimodal large models.

Computing Resources

Students may contact the instructor to request GPU resources. Each team should use fewer than 16 A100 GPUs (based on the global assignment).