Ramesh Arvind Naagarajan
PhD researcher at the Chair of Automatic Control and System Dynamics, TU Chemnitz. I work on explainable AI for control systems, using large language models as reasoning interfaces over physical processes, with mechanistic interpretability and causal discovery as the load-bearing tools. My current testbed is greenhouse climate control, a setting where the same controller has to satisfy a grower, an agronomist, and an optimiser at the same time.
I am part of the Automatic Control and System Dynamics group at TU Chemnitz, led by Prof. Dr.-Ing. Stefan Streif.
Research
Explainable model predictive control, causal reasoning, and language interfaces for engineers and operators.
Publications
Two journal papers in 2025 and a paper accepted at ICML 2026.
Writing
Notes on LLMs, interpretability, and the messy middle between control theory and machine learning.
CV
Background, experience, and the path here. PDF and web versions.