Codes for some of the completed projects:

  1. Data-driven Minimum Energy Control for Stochastic Linear Systems: Simulations for the paper Data-driven Minimum Energy Control for Stochastic Linear Systems, work done with Prof. Vaibhav Katewa at IISc. (Currently Private)

  2. Improving Reasoning in Multi-agent LLMs: A new two player game to understand collaboration among LLM agents with a new test metric to measure agent performance and suggestions for prompt variations that induce or discourage collaboration among the agents, work done with Priya Sarkar at IISc, advised by Prof. Aditya Gopalan.

  3. Consensus seeking in multi-agent systems: Project to demonstrate consensus dynamics in multi-agent systems using graph-based interactions prepared for Open day 2026 at IISc, work done with Samsad Alam at IISc.

  4. Bayesian approach to Robust RL: Code for Bayesian algorithms for robust RL, currently includes URBE, UBE and Vanilla DQN.

  5. MLP-blending: Multi-layer perceptron based blending limiter for hyperbolic PDEs (work done with Arpit Babbar at TIFR-CAM) (currently private).

  6. MECI-MLP: MLP based framework for minimum energy control input problem for linear dynamical systems with unknown dynamics (currently private).


Some other small codes that I have written (and think are useful). I try my best to update them frequently.

  1. Some algorithms to numerically solve Boundary value problems: Code.

  2. Some codes useful for understanding Tensorflow from basics: Code.

  3. Small codes to understand bandit algorithms: Code (More to come soon).

  4. CS-50 codes: Code (These were developed for CS-50 MOOC during my 1st year of UG. May not be very useful but I found the exercises useful and interesting).

  5. Pattern recoginition and neural networks course codes: .