Scholar
Amit Sharma
Google Scholar ID: CXgQufgAAAAJ
Principal Researcher, Microsoft Research
Citations & Impact
Citations
7,157
H-index
33
i10-index
80
Publications
20
Co-authors
26
list available
Publications
Resume (English only)
Academic Achievements
- Recipient of the NASSCOM AI Gamechangers Award (2023–24)
- Notable Top-25% paper at ICLR 2023 on out-of-distribution generalization
- Appointed Action Editor for Transactions on Machine Learning Research (TMLR)
- Invited speaker at NeurIPS 2024, AICPM 2023, and University of Cambridge workshops
- Published 'Causal reasoning and large language models: Opening a new frontier for causality' in TMLR (2024)
- Published 'Explaining machine learning classifiers through diverse counterfactual explanations' at FAccT 2020
Background
- Machine learning researcher focused on improving the reasoning abilities of AI systems
- Combines the generalizable capabilities of large language models (LLMs) with the principled nature of causal reasoning
- Works on applying causal AI to scientific discovery in medicine, climate science, and engineering
- Develops open-source tools like DoWhy and DiCE to enhance AI model reliability through causality
- Currently excited about 'Axiomatic Training', a framework for real-time correction of LLM outputs
