Aniket Das
PHD Student at Stanford University
Kanpur, Uttar Pradesh, United States
Summary
🤩
Rockstar🎓
Top SchoolAniket Das is a PhD student in computer science at Stanford’s Theory Group with eight years of research and engineering experience spanning ML, probability, and dynamical systems. He was a pre-doctoral researcher at DeepMind working on Markov chains, interacting particle systems, and algorithmic statistics, and has interned at Max Planck and TIFR on minimax optimization and PAC learning in MDPs. Trained in electrical engineering and mathematics at IIT Kanpur (with an exchange at Aalto), he combines rigorous theoretical depth with practical implementation skills. On GitHub he has contributed ML-focused tooling—implementing a range of GAN loss functions in the popular torchgan framework—and low-level backend work for a Python-to-Java transpiler, showing comfort across stacks. His profile reflects a knack for translating theoretical ideas into tested code and reproducible research. Colleagues can expect a researcher-engineer who moves fluidly between proofs, stochastic modeling, and production-quality implementations.
8 years of coding experience
2 years of employment as a software developer
Academic Exchange, School of Science, 4.8 / 5.0, Academic Exchange, School of Science, 4.8 / 5.0 at Aalto UniversityIndian Institute of Technology Kanpur
Github Skills (14)
javas10
pytorch10
machine-learning10
loss-functions10
deep-learning10
transcode10
python10
generative-adversarial-network10
transpiler10
java10
set-operations9
object-oriented-programming9
computer-vision9
testing8
Programming languages (2)
Github contributions (5)
torchgan/torchgan
Sep 2018 - Jan 2019
Research Framework for easy and efficient training of GANs based on Pytorch
Role in this project:
ML Engineer
ML EngineerContributions:33 commits, 31 PRs, 66 pushes in 4 months
Contributions summary:Aniket primarily contributed to the development of various loss functions for Generative Adversarial Networks (GANs) within the PyTorch framework. They implemented minimax, Wasserstein, and Least Squares GAN losses, including variations such as gradient penalty and label smoothing. Additionally, they added auxiliary classifier and feature matching losses, expanding the functionality of the `torchgan` framework.
pytorchpythondeep-learningtorchcomputer-vision
beeware/voc
May 2018 - May 2018
A transpiler that converts Python code into Java bytecode
Role in this project:Backend Developer
Backend DeveloperContributions:25 commits, 2 PRs, 2 comments in 5 days
Contributions summary:Aniket implemented and tested the `dict_keys` class, a core component for accessing dictionary keys in the Java-based transpilation of Python code. They added functionality to support set operations such as intersection, union, and difference on `dict_keys` objects. Subsequent commits focused on code style improvements, fixing indentation, and addressing Java linting issues. Further commits added implementation for `dict.values()` and `dict.items()` functionalities.
python-codepythonjava-bytecodebytecodecompiler
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.