A G
Portrait of Andrey Gizdov

About

Hi. I do research on compute-efficient and cognitively inspired multimodal models. I've been fortunate to work with Tomer Ullman at Harvard, Boris Katz at MIT, and Daniel Harari at the Weizmann Institute. I dropped my PhD to start something new.

No one got anywhere alone. A special thanks is due to:

Research

  1. 2025

    Towards aligned body representations in vision models

    Gizdov, Procopio, Li, Harari, Ullman

    @xai4science@aaai

  2. 2025

    Time-Series at the Edge: Tiny Separable CNNs for Wearable Gait Detection and Optimal Sensor Placement

    Procopio, Esposito, Raggiunto, Gizdov, Belli, Pierleoni

    @ai4ts@aaai

  3. 2025

    Seeing more with less: human-like representations in vision models

    Gizdov, Ullman, Harari

    spotlight@cvpr

  4. 2024

    Variable resolution improves VQA under a limited pixel budget

    Gizdov, Ullman, Harari

    @hcv@eccv

  5. 2018

    Digital cranial suture analysis with application in age estimation

    Gizdov, Nikolova, Toneva, Harizanov, Georgiev

    @isef

  6. Aligning vision models with human physical reasoning

    Gizdov, Procopio, Caputi, Ivanov, Li, Harari, Ullman

    @neuripsin review

  7. How predictable is AI progress?

    Mayo, Alabdulkareem, Shaw, Conwell, Gizdov, Katz, Barbu, Cheung

    in progress

Things I did

… etc.

Sometimes I write

Books I've been influenced by

… and many more.

Adventures

I like adventures and hard things.

Recently, I…