About me

Hi, I’m David! I’m a research scientist at Apple. I like training general-purpose multimodal models using simple yet scalable methods :)

Recent work

• Language Models Improve When Pretraining Data Matches Target Tasks
  David Mizrahi, Anders Boesen Lindbo Larsen, Jesse Allardice, Suzie Petryk, Yuri Gorokhov, Jeffrey Li, Alex Fang, Josh Gardner, Tom Gunter, Afshin Dehghan
  arXiv, 2025 • Paper

• FlexTok: Resampling Images into 1D Token Sequences of Flexible Length
  Roman Bachmann*, Jesse Allardice*, David Mizrahi*, Enrico Fini, Oğuzhan Fatih Kar, Elmira Amirloo, Alaeeldin El-Nouby, Amir Zamir, Afshin Dehghan
  ICML, 2025 • Project Page • Paper

• 4M-21: An Any-to-Any Vision Model for Tens of Tasks and Modalities
  Roman Bachmann*, Oğuzhan Fatih Kar*, David Mizrahi*, Ali Garjani, Mingfei Gao, David Griffiths, Jiaming Hu, Afshin Dehghan, Amir Zamir
  NeurIPS, 2024 • Project Page • Paper • Code

• 4M: Massively Multimodal Masked Modeling
  David Mizrahi*, Roman Bachmann*, Oğuzhan Fatih Kar, Teresa Yeo, Mingfei Gao, Afshin Dehghan, Amir Zamir
  NeurIPS, 2023 [Spotlight] • Project Page • Paper • OpenReview • Code

• MultiMAE: Multi-modal Multi-task Masked Autoencoders
  Roman Bachmann*, David Mizrahi*, Andrei Atanov, Amir Zamir
  ECCV, 2022 • Project Page • Paper • Code

* Equal Contribution