Gautham Bellamkonda

Caution: may spontaneously start talking about math and algorithms

I am currently working as an Algorithms Research Engineer at KLA. I graduated from IIT Hyderabad with a B.Tech in Computer Science and Engineering in 2024, where I had the pleasure to work with Prof. Karteek Sreenivasaiah on Quantum Query Complexity and Prof. Vineeth Balasubramanian on Logic Enhanced CBMs.

I have a strong passion for theoretical computer science and its applications to machine learning, with a focus on understanding why ML algorithms behave the way they do, what kinds of guarantees they can offer under various settings, and how we can leverage these insights to design better algorithms. I also enjoy exploring topics adjacent to these areas, such as Quantum Computing, Algorithmic Game Theory, Optimization theory, and Reinforcement Learning. At the end of the day, I’m motivated by clarity of thought, elegance of ideas, and the fun of problem-solving.

During high school, I was fortunate to be part of the International Mathematical Olympiad Training Camps in 2019 and 2020, after having qualified the Indian National Mathematical Olympiads in those two years. This experience ignited my love for problem-solving and deepened my appreciation for the beauty of mathematics. If you need any guidance or resources, feel free to reach out to me.

Outside of work and research, I enjoy sneaking into (theory) seminars on topics I know nothing about, over-analyzing movies, or just going for long walks on the beach.

Recent News

Dec 15, 2025	Going to attend ACM Winter School on Game Theory @ IIT Ropar
Nov 15, 2025	Our Paper “LogicCBMs: Using Logic to Enhance Concept-Based Learning” is accepted at WACV 2026!
Nov 07, 2025	Finished the course “Computability and Complexity” under Prof. Jayalal Sarma at IITM, while working at KLA. I am an IITM alum now!
Aug 14, 2025	Attended DSPOM 2025 held at ICTS-TIFR
Jul 03, 2025	Attended BIRS SAVI Workshop at CMI

Selected Publications

LogicCBMs

LogicCBMs: Logic-Enhanced Concept-Based Learning

Deepika SN Vemuri, Gautham Bellamkonda, Aditya Pola, and 1 more author

WACV 2026, Arizona, USA. More Information can be found here ,

Abs PDF

Concept Bottleneck Models (CBMs) provide a basis for semantic abstractions within a neural network architecture. Such models have primarily been seen through the lens of interpretability so far, wherein they offer transparency by inferring predictions as a linear combination of semantic concepts. However, a linear combination is inherently limiting. So we propose the enhancement of concept-based learning models through propositional logic. We introduce a logic module that is carefully designed to connect the learned concepts from CBMs through differentiable logic operations, such that our proposed LogicCBM can go beyond simple weighted combinations of concepts to leverage various logical operations to yield the final predictions, while maintaining end-to-end learnability. Composing concepts using a set of logic operators enables the model to capture inter-concept relations, while simultaneously improving the expressivity of the model in terms of logic operations. Our empirical studies on well-known benchmarks and synthetic datasets demonstrate that these models have better accuracy, perform effective interventions and are highly interpretable.