Past -> runwayml, Aleph-Alpha
Present -> Brains and computers @ murtylab
Science
- TopoNets - Brain-like topographic structure in vision & language models (ICLR 2025 Spotlight)
- AtMan - Memory efficient explainability method for anything with attention (NeurIPS 2023 + featured in Scientific American)
- DORA - Cluster and explore outlier neurons in vision models (TMLR, ICLR trustworthy ML workshop)
Contact: Telegram or email: [email protected]