Mathematics and Artificial Intelligence

Cheenta research school on Math and AI is focused on three key areas: AI assisted Mathematical Research, Mathematics assisted AI Research, and AI Resource Development for Teaching Mathematical Sciences.
LEARN MORE

How Artificial Intelligence Can Assist Mathematicians?

Researchers at Cheenta are working on research projects that explore how tools like LEAN 4, Isabelle and evolutionary algorithms can assist frontier research in mathematics.
Apply for research internship

Expository Paper: Exploring Hyperbolic Groups with Elementary Geometry and Genetic Algorithms 

This exposition presents an elementary geometric journey into hyperbolic geometry and group dynamics. Beginning with classical ideas such as cross ratio and inversion in a circle, we intuitively explore the hyperbolic metric on the unit disk using transformation geometry. We then study distance-preserving maps of the disk and use a ping–pong argument to construct a free group of isometries. By iterating these transformations, we observe the emergence of a fractal limit set structure resembling a Cantor set. Finally, we illustrate how simple evolutionary algorithms can computationally explore this limit set, offering a modern experimental perspective on classical geometric group theory.
Download Paper

Expository Paper: Is Math Curvy? Is Math Fat?

This exposition investigates the intrinsic geometric structure of human-formalized mathematical knowledge by embedding the Lean 4 Mathlib library into non-Euclidean manifolds. Using a dataset comprising approximately 440,000 theorems and millions of logical dependencies, we employ Hyperbolic Neural Networks (HNNs) and discrete Ollivier-Ricci curvature analysis to characterize the "shape" of the mathematical universe.

Our empirical results reveal a dual scale geometry: while the globally the graph of theorems is tree-like (Gromov-hyperbolic with $\delta=1$ ), locally it exhibits dense clustering (Jaccard-overlap proxy $\kappa \approx 0.022$ ). This tree of cliques reveals an intrinsic structure of the world of formalized mathematics.

We demonstrate the utility of this geometric duality through one toy example of AI-Mathematics collaboration: Manifold Extrapolation, where automated conjectures are generated by extending the manifold into peripheral leaf nodes.
Download Paper

How Artificial Intelligence can Assist Educators in Mathematical Science?

Research Workers at Cheenta are exploring avenues in behavioral psychology, pedagogical principle and building tools for educators in mathematical science.  
Apply for Research Internship

Randomized Control Trial on How Artificial Intelligence can Improve Human Collaboration in Virtual Classroom

In this experiment, we used three tests with same set of problems. The first test used the traditional format. The second test used an AI chatbot for discussion on each problem with the test - taker. In the third format of the test, AI is used to pair test-takers and induce a discussion between them before problems are attempted. We record the outcome of the experiment.
Researchers  at Cheenta
Dr. Ashani Dasgupta

Dr. Ashani Dasgupta

Dr. Ashani Dasgupta specializes in geometric group theory and low-dimensional topology. His research includes relative hyperbolicity, embedding problems, Morse conditions, local connectedness of boundaries for relatively hyperbolic groups, and connectedness properties of Bowditch boundaries.

Learn More
Srijit Mukherjee

Srijit Mukherjee

Srijit Mukherjee is a PhD student in the EECS Department at Pennsylvania State University (after studies in Statistics at Indian Statistical Institute). He builds interpretable AI solutions for biomedical imaging and diseases, collaborating with doctors and engineers at Harvard and Yale. His work explores expressing "mind, information, and intelligence" through mathematics and engineering, with applications in generative AI for problem-solving and novel problem creation.

Learn More
Arka Banerjee

Dr. Arka Banerjee

Arka Banerjee (Postdoc at Auburn University) focuses on metric geometry, geometric group theory, coarse geometry, coarse cohomology, coarse embeddings, group actions on CAT(0) cube complexes, coarse Poincaré duality spaces, and topological methods in geometry and topology.

Learn More
Ragunath JV

Ragunath JV

Raghunath is a PhD applicant in Pure Mathematics, with a focus on Topology, Group Theory, and allied areas. He love discussing Math Olympiad problems, particularly strategical game-theory, geometry, combinatorics problems.

Learn More
Swarnabja Bhaumick

Swarnabja Bhaumick

Swarnabja Bhaumick works in computer vision and machine learning, specializing in event and activity recognition in video surveillance for cyber-physical systems. His research uses deep learning methods (hybrid CNN-RNN architectures, graph-based key-frame selection, multi-tier feature fusion) to handle unconstrained videos from drones/CCTVs and achieve robust performance on benchmarks like UCF-101, HMDB, and CCV.

Learn More
Shayeef Murshid

Shayeef Murshid

Shayeef Murshid (Indian Statistical Institute) researches quantum computing and quantum information theory, with emphasis on quantum state discrimination, nonlocality, and the intersection with quantum cryptography — including secure encryption protocols and certified data deletion mechanisms.

Learn More
Shayeef Murshid

Dr. Sourayan Banerjee

Sourayan Banerjee is a PhD holder from IISER Bhopal and a Postdoctoral scholar at IIT Kanpur. His research focuses on Algebraic K-theory, exploring the intersections of algebra and topology.

Learn More
© 2010 - 2025, Cheenta Academy. All rights reserved.
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram