Anant Gupta

About Me

Hi! I’m Anant Gupta, a first-year M.S. student in Computer Science at Georgia Tech, where I also earned my B.S. in Computer Science and Mathematics. I’m interested in continual learning and meta-learning, and more broadly in building systems that learn, adapt, and teach like humans. I’m especially fascinated by how models can learn efficiently from limited data, generalize across domains, and reason about their own uncertainty and mistakes.

Before my M.S., I worked on chaotic dynamical systems and large-scale dataset generation. I’m also passionate about teaching and have been a TA for Automata and Complexity (CS 4510) for four semesters—an experience that strongly shapes how I think about learning in both humans and machines.

Research Goals

I’m particularly interested in models that know what they don’t know—systems capable of identifying uncertainty, detecting flawed reasoning, and revising their beliefs through reflection and interaction.

• Meta-learning: How models can acquire learning strategies that generalize across tasks, enabling fast adaptation, inductive bias discovery, and reuse of prior knowledge.

• Continual learning: Developing methods that allow models to learn over time without catastrophic forgetting, retaining and refining past knowledge while adapting to new tasks.

• LLMs as teachers: Using explanation, tutoring, and self-critique as learning mechanisms, where teaching helps models identify gaps in reasoning and improve long-term understanding.