About

I am a PhD student at Bar-Ilan University, advised by Prof. Ido Dagan.

My research tackles the “hallucination” problem in Large Language Models. I develop methods to ensure grounded text generation is factual and easily verifiable. I also explore how agents plan at test-time and how we can better align text-to-image generation.

Previously, I spent over a decade in the industry as a Data Scientist, Software Engineer and Team Lead.

📌 Selected Research

LAQuer teaser

LAQuer: Localized Attribution Queries

ACL 2025 — We introduce a method for fine-grained attribution in content-grounded generation, allowing models to cite specific spans rather than full documents.

Paper Code

Attribute First teaser

Attribute First, then Generate

ACL 2024 — We propose a multi-step generation process that creates localized attribution: first select spans from the source, then create a plan based on these spans. Finally use the plan to generate the output.

Paper Code

📢 News

  • Nov 2025 PrefixNLI preprint is now available online.
  • Jun 2025 GenerationPrograms accepted to COLM '25!
  • Jun 2025 Released CLATTER preprint.
  • May 2025 LAQuer accepted to ACL '25 main conference.
  • Feb 2025 Make it Count accepted to CVPR '25.