Lj V. Miranda

Can you bootstrap your way from annotating data to training a frontier model in one thousand days?

I gave a talk at the Analytics & AI Association of the Philippines (AAP) on the topic of Philippine language model evaluation (Slides), specifically on FilBench.1 The purpose of the meetup was really to figure out (1) whether we can build a “Philippine LL…

Back in 2021, I decided to embark on a PhD program and pursue an academic research career.

At the end of 2024, I wrote about my desiderata for Filipino NLP.

I had an incredible time at the ACL 2025 Conference in Vienna.

Recently, we’ve witnessed language models become increasingly adept at using real-world tools through their function-calling capabilities.

Back when I started working in Filipino NLP, my standard approach in training models is to encode linguistics knowledge via meticulous data annotation, feature engineering, and extensive testing.

I was invited to give a talk to a graduate-level NLP class about my work on Filipino resources.

Preference data is a staple in the final step of the LLM training pipeline.

A few weeks ago, I held a guest lecture in the DSBA 6188: Text Mining and Information Retrieval Class at UNC Charlotte on using large language models (LLMs) for annotation.