Scholar
Sujay Sanghavi
Google Scholar ID: O-DazBUAAAAJ
Professor, Electrical and Computer Engineering, University of Texas, Austin
Citations & Impact
Citations
5,582
H-index
36
i10-index
75
Publications
20
Co-authors
42
list available
Publications
Resume (English only)
Academic Achievements
- ICML 2025 Papers:
- - Upweighting Easy Samples in Fine-Tuning Mitigates Forgetting
- - Learning Mixtures of Experts with EM: A Mirror Descent Perspective
- - Retraining with Predicted Hard Labels Provably Increases Model Accuracy
- - Geometric Median Matching for Robust k-Subset Selection from Noisy Data
- ICLR 2025 Papers:
- - Enhancing Language Model Agents using Diversity of Thoughts
- - Infilling Score: A Pretraining Data Detection Algorithm for Large Language Models
- Graduated PhD students: Rudrajit Das (headed to Google Research) and Anish Acharya (headed to AWS)
- Completed MS degrees: Atula Tejaswi (continuing at UT for a PhD) and Vijay Lingam (headed to Amazon Q)
- Invited talk in Workshop on Theoretical Perspectives on LLMs (March 2025)
Research Experience
- Since 2019, he has been a Principal Research Scientist and then an Amazon Scholar, working in Amazon’s Search org; Visiting Scientist at Google Research; Senior quant and founding member of an algorithmic trading team at the hedge fund Engineers Gate.
Background
- Research interests are in machine learning, with a current focus on understanding and improving the architecture and training of large-scale models for language and representation learning. Also broadly interested in understanding machine learning from a theoretical perspective.
