cv

My academic journey, in broad strokes.

General Information

Full Name Saksham Jain
Languages Hindi, English, Japanese, French (Beginner)

Education

  • 2027 (Expected)
    PhD, Statistics
    University of Washington-Seattle, USA
    • Conditional distributional treatment effects - doubly robust estimation and testing
    • Longitudinal causal inference with multiple continuous exposures - an application to proteomics (Working Paper)
  • 2022
    MS, ECE (ML Concentration)
    Duke University, USA
    • Evaluating pre-trial programs using interpretable ML matching methods for causal inference (Master's Project)
    • Closing the domain gap - blended synthetic imagery for climate object detection
  • 2019
    BE, Robotics & Automation
    University of Delhi, India
    • Synthetic data augmentation for surface defect detection and classification using deep learning (Bachelor's Thesis)

(Select) Experience

  • Jun'25 - Sep'25
    Research Intern
    RIKEN-AIP and the University of Tokyo
    • State evolution for gradient descent in finite-width multi-layer neural networks
  • Jun'24 - Sep'24
    Applied Scientist II Intern
    Amazon Science
    • LLMs for demand forecasting
  • Jan'20 - Aug'20
    Research Assistant
    Indian Institute of Science Bangalore
    • Training-data-free distillation
    • Smart super-resolution for sattelite imagery
  • Jun'19 - Jan'20
    Machine Learning Engineer
    Million Sparks Foundation
    • Recommendation systems

Academic Interests

  • Causal Inference
    • Distributional treatment effects
    • Counterfactual sampling from generative models
  • Deep Learning
    • Precise gradient descent dynamics, and applications
  • Interactive ML
    • Adaptive experimental design
    • Dynamic treatment regimes