This is the official implementation for "Mixture of Mini Experts: Overcoming the Linear Layer Bottleneck in Multiple Instance Learning". We are currently preparing the codebase for public release. Please check back soon.
If you find this work useful in your research, please cite:
@inproceedings{shao2026mixture,
title={Mixture of Mini Experts: Overcoming the Linear Layer Bottleneck in Multiple Instance Learning},
author={Shao, Daniel and Runevic, Joel and Chen, Richard J. and Williamson, Drew F.K. and Kim, Ahrong and Song, Andrew H. and Mahmood, Faisal},
booktitle={Fourteenth International Conference on Learning Representations (ICLR)},
year={2026}
}
\```