Noam Slonim

Also known as נעם סלונים

Noam Slonim is an AI researcher whose work spans information theory, computational argumentation and language models. He completed a PhD at Hebrew University's Interdisciplinary Center for Neural Computation in 2002 and spent four years as a Princeton postdoc in physics and genomics. At IBM Research, he led the Debating Technologies Grand Challenge, coauthored a 2021 paper about Project Debater, published a Key Point Analysis demo with colleagues and studied zero-shot text classification with large language models. Since September 2025, he is a research scientist at Google Research Israel, where he works on Gemini and coauthors a 2026 paper about Google Earth AI's Geospatial Reasoning agent.

Now at: Google
Works: 177
Citations: 5,459
h-index: 32
Badges: Bridge

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

2002

A Hebrew University PhD

He completes a PhD at the Hebrew University's Interdisciplinary Center for Neural Computation in 2002, working in its Machine Learning lab.
2003–2006

Princeton postdoctoral research

From 2003 to 2006, he spends four years as a postdoctoral researcher in Princeton's Department of Physics and the Lewis-Sigler Institute for Integrative Genomics. There, he develops techniques for analyzing genomic data.
2006

Multivariate Information Bottleneck

In 2006, he coauthors a general framework that extends the Information Bottleneck method to multiple interrelated data partitions.
2007–2011

IBM's Debating Technologies challenge

He joins the IBM Haifa Research Lab as a research staff member in 2007. In 2011, he becomes technical lead of IBM's Debating Technologies Grand Challenge and serves as its principal investigator.
2021

An autonomous debating system

In 2021, he coauthors a Nature paper that presents Project Debater as an autonomous system for competitive debate with humans. The paper describes its architecture and evaluates its performance, including its public debut against three expert debaters.
November 2021

Key Point Analysis demo

In November 2021, he and colleagues publish an online demo of Key Point Analysis, which identifies main points and their prevalence in opinionated text. They also organize a shared task at the ArgMining workshop.
December 2023

Zero-shot text classification

In December 2023, he coauthors a study of zero-shot topical text classification with large language models. The study compares models on TTC23, a benchmark of 23 publicly available datasets across domains and styles.
Since September 2025

Google Research and Earth AI

He joins Google Research Israel as a research scientist in September 2025. He works on efforts to enhance Gemini and advance AI reasoning and language understanding. In October 2026, he coauthors a paper about Google Earth AI's Geospatial Reasoning agent, a Gemini-powered system that combines geospatial data sources, foundation models and tools to answer questions about our planet.