Hongyuan Mei
Scholar

Hongyuan Mei

Google Scholar ID: g_zaiVIAAAAJ
Google DeepMind, TTIC, JHU, UChicago
ReasoningLarge Language ModelsNatural Language UnderstandingMachine Learning
Citations & Impact
All-time
Citations
2,032
 
H-index
19
 
i10-index
24
 
Publications
20
 
Co-authors
12
list available
Resume (English only)
Academic Achievements
  • - Release of Grok 4 and Grok 4 Heavy, achieving SOTA on Humanity’s Last Exam.
  • - CaseSumm accepted to NAACL 2025.
  • - Mango accepted to COLM 2024.
  • - Statler accepted to ICRA 2024.
  • - EasyTPP accepted to ICLR 2024.
Research Experience
  • - Member of Technical Staff at xAI, responsible for training Grok to be a knowledgeable reasoner and contributing to the development of Grok 4 Heavy.
  • - Senior Research Scientist at Google DeepMind before joining xAI.
  • - Research Assistant Professor at Toyota Technological Institute at Chicago (TTIC), leading the River Lab in research on AI reasoning and planning.
Education
  • PhD in Computer Science from Johns Hopkins University, advised by Jason Eisner. Developed the neural Hawkes process, widely used in event sequence modeling and prediction.
Background
  • Research Interest: Training computers to reason; Professional Field: AI reasoning and planning; Bio: Currently working at xAI, dedicated to developing the knowledgeable reasoning system Grok.