Chandan Tankala
Researcher in Mathematical Foundations of AI
I recently completed my Ph.D. in Mathematics at the University of Oregon under the supervision of Dr. Krishnakumar Balasubramanian, where I developed theoretical foundations for modern machine learning algorithms. Prior to my Ph.D., I earned my M.S. degree from Texas A&M University, advised by Dr. Robert Weiss.
My research lies at the intersection of artificial intelligence, statistics, and theoretical computer science. During my Ph.D., I focused on developing theoretical foundations for sampling and optimization algorithms. Currently, I am expanding my research portfolio to encompass generative AI, reinforcement learning, and practical applications of AI algorithms, while continuing to advance my work in algorithmic theory.
I actively collaborate with researchers across leading institutions, including Dr. Dheeraj Nagaraj at Google DeepMind, Dr. Anant Raj at the Indian Institute of Science, and Dr. Quan Zhou at Texas A&M University, working on sampling and optimization problems.
Research Areas
Research Highlights
Working with Dr. Dheeraj Nagaraj (Google DeepMind) on mean field optimization via interacting particle systems, developing novel stochastic algorithms that reduce the number of required particles from exponential to polynomial.
Working with Dr. Dheeraj Nagaraj (Google DeepMind), developed novel stochastic algorithms for mean-field optimization that go beyond traditional propagation of chaos methods, establishing rigorous convergence guarantees.
Working with Dr. Krishnakumar Balasubramanian (UC Davis), we extended dense associative memories from Euclidean space to the Wasserstein manifold of probability distributions, enabling principled storage and retrieval of Gaussian measures with applications to generative modeling and uncertainty-aware computation. This framework bridged classical memory architectures with modern generative AI by treating entire distributions as computational units. This work has been submitted to ICLR 2026.