Jathushan Rajasegaran
PHD Student at University of California, Berkeley
Sri Lanka
Summary
👤
Senior🎓
Top SchoolJathushan Rajasegaran is a PhD student at UC Berkeley with a decade of experience focused on deep learning architectures, optimization, and addressing practical weaknesses of neural networks. His research blends statistical machine learning—especially spectral methods for latent variable models—with lifelong and meta-learning, reflecting a rare mix of theoretical and applied interests. Prior roles at the Inception Institute of AI, Singapore Management University, and CSIRO Data61 show consistent research-engineering contributions across academia and industry. A department gold medalist in Electronics and Telecommunications from the University of Moratuwa, he brings strong mathematical foundations to complex ML problems. Based in Sri Lanka, he combines global research experience with hands-on R&D from IoT prototyping to large-scale ML systems. Colleagues describe him as someone who bridges rigorous theory and reproducible engineering, often uncovering insights that improve both model performance and trainability.
10 years of coding experience
2 years of employment as a software developer
Bachelor of Science - BS, Electronics and Telecommunication, Department Gold medalist (CGPA 4.1/4.2), Bachelor of Science - BS, Electronics and Telecommunication, Department Gold medalist (CGPA 4.1/4.2) at University of MoratuwaAdvanced Level, Mathematics, District 1st and Island level 3rd., Advanced Level, Mathematics, District 1st and Island level 3rd. at Jaffna hindu College
English, Tamil, Sinhalese
Github Skills (36)
meta9
meta-learning9
supervised-learning9
representations9
transfer-learning8
few-shot-learning8
incremental-learning8
pytorch8
deep-learning7
3d-reconstruction7
computer-vision6
maml6
vector5
artificial-intelligence5
benchmark5
Programming languages (3)
Github contributions (5)
brjathu/iTAML
Nov 2019 - Apr 2020
Official implementation of "iTAML : An Incremental Task-Agnostic Meta-learning Approach". CVPR 2020
Contributions:35 commits, 5 pushes, 21 comments in 5 months
pytorchsupervised-learningincrementalmeta-learningdeep-learning
brjathu/SKD
Jun 2020 - Apr 2022
SKD : Self-supervised Knowledge Distillation for Few-shot Learning
Contributions:9 commits, 2 pushes, 1 comment in 1 year 10 months
pytorchdistillationsupervisedknowledge-distillationmeta-learning
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.