Siddharth

I am a CS undergrad @ SRMIST interested in deep AI work. Most of what I do involves implementing and probing research ideas. I am drawn to architecting systems that move toward general intelligence.

I enjoy tackling scientific and applied problems.

Email  /  GitHub  /  LinkedIn  /  CV

Selected Work

A tool that converts research papers into structured, explorable representations to support deep technical reading and understanding.

Code

DreamerV1 world model implementation on Gridworld tasks, based on Dream to Control (Hafner et al., 2019).

Code

Implemented and trained Deep Q-Networks on the Atari game Breakout.

Code

A system to flag anomalous tourists for safety, using synthetically generated data and supervised learning.

Code

Misc

Article — Planning by Dreaming: Why DreamerV1 Breaks in Simple Worlds

Coursework — Stanford CS229