Undergraduate work in dynamical systems, atmospheric data, mathematical writing, and statistics, newest first.
Dynamical Systems
Independent Study
Aug 2026 – Present
Parameterization of Lorenz-96 and Testing Online Stability Against Training Data Volume
How much training data does a neural network need to parameterize the unresolved subgrid coupling $U_k$ without drift or instability when it is integrated online in the reduced (slow-variable only) Lorenz-96 model, as opposed to merely being accurate offline?
Running a data-volume sweep across 10 training-data volumes for two models with different activation functions, 20 models in all.
Comparing offline fit ($R^2$, MSE) against long-horizon online stability for each model.
Mathematical Writing
Open Source
Oct 2026 – Present
Mathematical Methods Handbook
A website-book of the mathematics physicists and applied scientists use. It is split into small units you can read on their own, select, and download as a PDF.
Written from a library of 19 textbooks, with Arfken, Weber & Harris, Mathematical Methods for Physicists (7th ed.) as the spine.
Every block of every unit is labeled Original, Adapted (with the book and pages it follows), or Quoted.
Software checks each unit against the full text of the source books before it is published.
Monsoon Dynamics
IDEAS-TIH, ISI Kolkata
May 2026 – Jul 2026
Impact of ENSO on Rainfall Variability Across Major Kharif Crop Growing States
Research project and training program with IDEAS-TIH, ISI Kolkata. I built an interpolation method to obtain higher-resolution rainfall data than the IMD portal provides, automated the per-state workflow, and classified rainfall variability using ENSO sensitivity indices.
Carried out exploratory data analysis of IMD gridded daily rainfall data.
Developed an interpolation method to obtain rainfall at a higher resolution than the IMD portal provides.
Automated the Google Sheets workflow with Python and the Google Cloud APIs so the analysis runs per state.
Developed a metric for rainfall variability and classified states using the ENSO Sensitivity Index and related measures.
Science Education
ProblemSolvers Lab, IISER Pune
Jun 2026 – Jul 2026
Patterns of Reasoning and Factors of Growth
Research project with the ProblemSolvers Lab, Department of Science Education, IISER Pune, on how students reason through pattern-recognition problems.
Reviewed earlier work on the topic.
Cleaned and organized participant data over several rounds.
Created a new conceptual framework for the reasoning patterns seen in pattern-recognition problems.
Documented the existing rubric used to measure performance.
Analysis
Reading Project, IISER Pune
Dec 2025 – Feb 2026
Metric Spaces
A reading project through Mícheál Ó Searcóid's Metric Spaces, working problems at every level of difficulty.
Worked problems on coding metrics such as the Hamming and Levenshtein metrics, the $p$-adic metric, and normed linear spaces.
Applied Statistics
UC Berkeley REYES
Jun 2025 – Oct 2025
Statistical Instrument on Circular Economy
Part of the UC Berkeley REYES mentorship program: measuring public awareness of the circular economy and mapping the current research on it.
Ran a survey targeting more than 1,000 responses on awareness of the circular economy.
Collaborated on a literature review of more than 300 articles.
Analyzed qualitative and quantitative research questions from the literature review.
Analyzed the survey data with hypothesis testing, ANOVA, and regression.