I am a Ph.D. student in Computer Science at Stanford University, working on NLP and ML.

Nowadays I mostly think about how to better decouple knowledge from capabilities in language models. Can we build smaller models that don't memorize half the internet, yet can reason and perform complex tasks? I am also interested in multimodal representaion learning, and in principled and effective methods to contexualize models with external information.

Before Stanford, I was working as an engineer at Meta.
I completed my B.Sc. and M.Sc. in Computer Science at Ben-Gurion University, where I worked on evaluation of tokenization algorithms for language models.

Feel free to reach out!

Selected Papers

• 
  Guided Query Refinement: Multimodal Hybrid Retrieval with Test-Time Optimization

  
  

  Omri Uzan, Asaf Yehudai, Roi Pony, Eyal Shnarch, Ariel Gera

  ICLR-26

  paper code project
• 
  CharBench: Evaluating the Role of Tokenization in Character-Level Tasks

  
  

  Omri Uzan, Yuval Pinter

  AAAI-26 Oral

  paper 🤗benchmark

• 
  Greed is All You Need: An Evaluation of Tokenizer Inference Methods

  
  

  Omri Uzan, Craig W. Schmidt, Chris Tanner, Yuval Pinter

  ACL-24 Oral Outstanding Paper Award Senior Area Chair Award

  paper code