Yuval Kirstain

About

I’m a research scientist at Google DeepMind, working on pretraining data.

I worked on Muse Image (Meta TBD), Movie Gen (Meta GenAI), Pick-a-Pic and PickScore (PhD), and Emu Edit (Meta FAIR).

My doctoral research at Tel Aviv University with Omer Levy spanned natural language processing and generative modeling.

Education

Ph.D. in Computer Science
Tel Aviv University · 2019–2024
Advisor: Omer Levy

B.Sc. in Computer Science, magna cum laude
Hebrew University of Jerusalem · 2016–2019

Industry Experience

• Research Scientist · Google DeepMind
  Oct 2026–Present
  Pretraining data.

• Research Scientist · Meta · TBD
  Jul 2025–Sep 2026
  Data curation and generation, model training, and evaluation.

• Research Scientist · Meta · GenAI
  Oct 2023–Jun 2025
  Movie Gen training and video editing.

• Research Scientist Intern · Meta · FAIR
  May 2023–Sep 2023
  Co-first author of Emu Edit.

• Research Scientist Intern · Meta · FAIR
  May 2022–Nov 2022
  Text-to-image generation.

• Research Scientist Intern · IBM Research
  Aug 2021–Oct 2021
  Task-oriented dialogue datasets and model training.

• Algorithm Developer · AI21 Labs
  Feb 2020–Mar 2021
  Large language model training and evaluation.

Publications

Google Scholar

1. Muse Image

   Worked on Meta’s image generation model.

   Meta · 2026 · Project release

   Project announcement

2. Movie Gen: A Cast of Media Foundation Models

   Adam Polyak et al. (including Yuval Kirstain)

   Technical report · 2024

   Paper

3. Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation

   Yuval Kirstain, Adam Polyak, Uriel Singer, Shahbuland Matiana, Joe Penna, Omer Levy

   NeurIPS 2023

   Paper/Code & models

4. Emu Edit: Precise Image Editing via Recognition and Generation Tasks

   Shelly Sheynin*, Adam Polyak*, Uriel Singer*, Yuval Kirstain*, Amit Zohar*, Oron Ashual, Devi Parikh, Yaniv Taigman

   CVPR 2024

   Paper/Project

5. VideoJAM: Joint Appearance-Motion Representations for Enhanced Motion Generation in Video Models

   Hila Chefer, Uriel Singer, Amit Zohar, Yuval Kirstain, Adam Polyak, Yaniv Taigman, Lior Wolf, Shelly Sheynin

   ICML 2025 · Oral

   Paper/Project

6. Few-Shot Question Answering by Pretraining Span Selection

   Ori Ram*, Yuval Kirstain*, Jonathan Berant, Amir Globerson, Omer Levy

   ACL-IJCNLP 2021

   Paper

7. A Few More Examples May Be Worth Billions of Parameters

   Yuval Kirstain, Patrick Lewis, Sebastian Riedel, Omer Levy

   Findings of EMNLP 2022

   Paper

8. Through-The-Mask: Mask-based Motion Trajectories for Image-to-Video Generation

   Guy Yariv, Yuval Kirstain, Amit Zohar, Shelly Sheynin, Yaniv Taigman, Yossi Adi, Sagie Benaim, Adam Polyak

   CVPR 2025

   Paper/Project

9. Video Editing via Factorized Diffusion Distillation

   Uriel Singer*, Amit Zohar*, Yuval Kirstain, Shelly Sheynin, Adam Polyak, Devi Parikh, Yaniv Taigman

   ECCV 2024 · Oral

   Paper/Project

10. Coreference Resolution without Span Representations

    Yuval Kirstain*, Ori Ram*, Omer Levy

    ACL-IJCNLP 2021

    Paper

11. X&Fuse: Fusing Visual Information in Text-to-Image Generation

    Yuval Kirstain, Omer Levy, Adam Polyak

    arXiv 2023

    Paper

* Equal contribution.

Contact

ykirstain@gmail.com

Find me on Twitter and LinkedIn.