Who Leads Machine Learning Research? A Bibliometric Portrait of 1,000 Top-Cited Researchers on Google Scholar

We introduce Scholar-ML-1000, the first publicly released dataset of the 1,000 most-cited researchers on Google Scholar who self-identify with the label machine learning, collected in June 2026. After removing four anomalous accounts (citation-inflated or off-topic), the clean dataset of 996 researchers spans a citation range of 40,809–1,118,561 (median 69,320; mean 101,379; Gini coefficient 0.37). Academia remains the dominant sector: 52.3% of researchers hold primarily academic affiliations, accounting for 52.3% of total citations, while pure-industry researchers (22.6%) account for 26.6%. At the company level, Google and Google DeepMind together list 209 researchers—more than Stanford, MIT, and UC Berkeley combined (74 researchers). Meta, OpenAI, and Microsoft each contribute 20–35 researchers; Anthropic and NVIDIA each contribute approximately 10. US-based researchers constitute 41.3% of the dataset; the remaining 58.7% are distributed across Europe, China/Hong Kong, Canada, and beyond. Research interests span well beyond the traditional ML core: particle physics (49 researchers), computational biology (50), and astrophysics (15) each form distinct communities that adopt ML as a primary tool. We release the dataset, collection scripts, and analysis code to support further science-of-science research on the global ML talent ecosystem.