Projects

ML-Track

  • Interpretability-Guided Unlearning in Audio
    [Undergraduate Thesis]: Interpretability-guided machine unlearning (MU) for audio models, using SAE-driven feature mining and LEACE concept edits, evaluated on speaker and speech recognition datasets across English, Hindi, and Spanish. Beats SoTA concept erasure and MU methods.
  • Mechanistic Analysis of Alignment Algorithms
    Mechanistic analysis of how alignment algorithms shape language model representations. Compared DPO, GRPO, KTO, PPO, and SimPO with activation probes, sparse autoencoders, and crosscoders.
  • "Fantastic Biases and Where to Find Them"
    Mechanistic interpretation of racial and gender bias in LLMs using linear probing, activation and attribution patching, and circuit tracing. Analyzed GPT-2 (Medium/Large) and GPT-Neo models on StereoSet and WinoGender datasets to identify where bias emerges in model representations.
  • 3D Scene Reconstruction
    End-to-end 2D to 3D scene reconstruction using classical, hybrid, and learning-based strategies including SfM, MVS, GANs, Neural Radiance Fields, and Gaussian Splattings.
  • FastSim with CERN
    Developing flow-based, diffeomorphic generative models for "Fast Shower Simulation (FastSim)" in High Energy Physics.

Core CS-Track

  • DQN-based Maze Solver on STM32 Microcontroller
    Built a maze environment for an RL agent to solve. The model was then effeciently deployed on an STM32 Cortex-M4 microcontroller post 4-bit quantization.
  • Distributed Program Execution
    Byzantine Fault Tolerance in Peer-to-Peer Networks. Built with OmNET++ Simulator.
  • Huffman Encoder
    File compression system using Huffman encoding algorithm.