All Posts

In the second part of this blog post series, we are going to explore new ways in which we can collect geolocated text (i.e. text tied to a geographic position on Earth).

Finding the pair index whose RoPE frequency makes N rotations over the sequence length.

The 6NP rule is a shortcut for estimating how many floating-point operations are needed to train a dense neural network.

A blog post where I try to give a rough estimate of my codex usage impact.

Walking our way through masked image modeling

A blog post to breakdown the maths behind attention pooling.

A new video from my YouTube channel.

A blog post where I am introducing a simple first approach to describe a given place on Earth using wikidata

A few words on why I am starting a youtube channel

A new video from my YouTube channel.

A new video from my YouTube channel.

How a Java Monopoly project in school set me on the path to computer science.

First review!

Analyzing OpenStreetMap keys and tags to surface what's relevant to landuse, from raw counts down to clustered key families.

A short look at why Hugging Face language tags can be useful while still requiring manual investigation.

Our Journal of Simulation paper on LLMs, agent-based models, and empathetic decision-making was selected as the Editor's Pick.

A look at our new research on using LLMs to turn complex ABM simulations into clear textual explanations.

Next month, I will be joining INRIA in Montpellier, within the EVERGREEN research team as a research engineer.

Using the perfect cup of chai to understand the fundamentals of Design of Experiments (DOE)

A breakdown of the How2Bench paper, which advocates rigor in benchmark development, with a focus on evaluation reliability and reproducibility.

Some advice and resources I have found helpful so far as a junior AI researcher.

A comparison between CLI and MCP approaches for giving AI agents capabilities to interact with systems.

A breakdown of the GeoReasoner paper, which leverages both linguistic and geospatial information to reason on geospatially grounded natural language.