Aritra Bandyopadhyay

ML Researcher, IIEST Shibpur

I work across mechanistic interpretability, computer vision, and information retrieval. This Jekyll version is intentionally designed to evoke the reading experience of a classical academic paper: serene, readable, and deeply focused on content.

Browse Publications Read Notes Curriculum Vitae

GitHub LinkedIn Google Scholar

Research Focus

The site comparison is practical: keep the same route shape and content inventory while letting each framework show its natural defaults. In Jekyll, that means collection-driven pages, a clean reading layout, and a strong bias toward publication and CV workflows.

Featured Publication

Differential Evolution Algorithm Based Hyper-Parameter Selection of Convolutional Neural Networks for Speech Command Recognition

Evolutionary Computation Theory and Applications (IJCCI / ECTA) — 2023

An evolutionary search approach for CNN hyper-parameters in speech command recognition.

Read Paper

Selected Open Source

Pandas

Flexible and powerful data analysis and manipulation library for Python.

Source Code

PyTorch

Deep learning framework for research code and production prototypes.

Source Code

Recent Notes

Come back later

October 24, 2023

A placeholder notebook-style post for future language model notes.

Read Entry