Anant Gupta

Atlanta GA, 30332

About Me

Hi, I’m Anant Gupta, a second-year M.S. student in Computer Science at Georgia Tech, where I also completed my B.S. in CS and Mathematics.

My core research focuses on continual learning and self-improving AI systems capable of learning like humans from limited data. To maximize the training signal from minimal data, my current work explores concepts of compositionality and self-play to build autonomous internal evaluation loops.

I’ve also been a TA for Automata and Complexity (CS 4510) for four semesters, which strongly influences how I think about learning in both humans and machines. This has led to my work on applying LLMs to educational technology, specifically building scalable benchmarks and systems for intelligent tutoring in high-complexity domains like algorithms and complexity theory.

Previously, I have also conducted research in chaotic dynamical systems and led large-scale financial dataset curation for LLM training and evaluation.

selected publications

ICLR 2026

Avoid Catastrophic Forgetting with Rank-1 Fisher from Diffusion Models

Zekun Wang^*, Anant Gupta^*, Zihan Dong, and 1 more author

In The Fourteenth International Conference on Learning Representations, 2026
Preprint

Self-Consolidating Language Models: Continual Knowledge Incorporation from Context

Zekun Wang^*, Anant Gupta^*, Zihan Dong, and 1 more author

2026
Preprint

Test-Time Compositional Generalization in Diffusion Models via Concept Discovery

Zekun Wang, Anant Gupta, Tianyi Zhu, and 1 more author

arXiv preprint arXiv:2605.07078, 2026
Preprint

Trust Region Continual Learning as an Implicit Meta-Learner

Zekun Wang^*, Anant Gupta^*, and Christopher J MacLellan

2026