Rohit Gandikota
Research Scientist Intern at Adobe
Boston, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolRohit Gandikota is a Ph.D. candidate at Northeastern University specializing in interpretable and controllable generative AI, with eight years of experience building and deploying neural networks at scale. His research blends causal interventions and concept erasure to align models with human values, producing methods that can permanently remove harmful or copyrighted concepts from diffusion models without retraining. Before his Ph.D. he pioneered data-hiding in generative models—work that won India’s Best Innovative Computer Science Thesis—and built the first deep generative model in production at ISRO, enabling real-time flood monitoring and dramatically accelerating satellite image production. He has contributed to top venues (ICCV, BMVC, IGARSS), interned at Adobe and Wayfair on user-facing generative controls, and collaborated with MIT and Technion on unified concept editing. Comfortable across research and production stacks (PyTorch, TensorFlow, GCP, Docker, Kubernetes), he aims to create an “atlas of AI” that maps model capabilities, relationships, and vulnerabilities to make systems safer and more transparent.
8 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD Interpretable Generative AI (Advisor: Prof. David Bau), Doctor of Philosophy - PhD Interpretable Generative AI (Advisor: Prof. David Bau) at Northeastern UniversityBachelor of Technology - BTech Electronics and Communications Engineering (Specialization in Avionics), Bachelor of Technology - BTech Electronics and Communications Engineering (Specialization in Avionics) at Indian Institute of Space Science and Technology
Github Skills (16)
pytorch8
audio8
flux8
stable-diffusion7
deep-learning7
diffusion-models6
image-generation5
diffusers4
machine-learning3
text2image3
jax3
text-processing2
artificial-intelligence2
nlp-machine-learning1
ai1
Programming languages (2)
Github contributions (5)
rohitgandikota/sliders
Nov 2023 - Mar 2025
Concept Sliders for Precise Control of Diffusion Models
Contributions:1 release, 2 PRs, 39 pushes in 1 year 4 months
rohitgandikota/hiding-audio-in-images
Nov 2019 - Dec 2021
Contributions:6 commits, 5 pushes, 1 branch in 2 years 1 month
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.