Scholar
Scott Lundberg
Google Scholar ID: ESRugcEAAAAJ
Microsoft AI
Citations & Impact
Citations
66,542
H-index
45
i10-index
56
Publications
20
Co-authors
3
list available
Publications
Resume (English only)
Academic Achievements
- Created Guidance: an open-source library for controlling LLMs, enabling structured generation and fast execution; used by OpenAI for JSON output mode
- Developed SHAP (SHapley Additive exPlanations): a widely adopted package for explaining any ML model using theoretically grounded SHAP values
- Built Prescience: an explainable ML system for real-time hypoxaemia risk prediction during surgery, improving anesthesiologist performance
- Proposed exact game-theoretic explanations for tree ensembles, enabling polynomial-time computation of Shapley values
- Developed ChromNet: inferred human chromatin network from ENCODE ChIP-seq data
- Introduced a unified framework (SHAP) for interpreting model predictions, reconciling multiple explanation methods
