Your Photo

Lara Marinov Hello! Buna! Guten Tag! Bonjour ! Buongiorno! Zdravo! Ahoj!

/’lɑːɹə m'e:ɹɪnˌɔːv/

I will be starting as a Ph.D. student at UT Austin in Fall 2026!

I graduated in Fall 2025 with my B.S. in Computer Science with concentrations in Software Engineering and Language Technologies from Carnegie Mellon University. I was on an exchange semester at École Polytechnique Fédérale de Lausanne (EPFL) in Spring 2025.

I’m passionate about solving real-world problems that involve both programming and human languages. My interests lie at the intersection of NLP and software engineering, where AI can improve the way we build and interact with software.

GitHub LinkedIn Google Scholar Email

Research

My research interests span various topics in NLP, software engineering, and linguistics.

Paper thumbnail

Discovering Functionally Selective Brain Regions with a Deep Topographic Multimodal Model

Badr AlKhamissi, Johannes Mehrer, Lara Marinov, Ahmed Abdelaal, Abdulkadir Gokce, Martin Schrimpf; 2026

PDF
Paper thumbnail

Not-Just-Scaling Laws: Towards a Better Understanding of the Downstream Impact of Language Model Design Decisions

Emmy Liu, Amanda Bertsch, Lintang Sutawika, Lindia Tjuatja, Patrick Fernandes, Lara Marinov, Michael Chen, Shreya Singhal, Carolin Lawrence, Aditi Raghunathan, Kiril Gashteovski, Graham Neubig; 2025

PDF

Projects

Here's a selection of projects I've done recently.

Poster thumbnail

User Retention Analysis

We analyzed behavioral data from GoGymi, a digital study app aimed at middle school students, to identify key early signals of long-term user retention. We focused on three main research questions and used a number of ML models to analyze the data. Check out our poster for more details on our analysis. This project was done as team of three for the course "Machine Learning for Behavioral Data" at EPFL.

Poster
Paper thumbnail

Scene4M

Scene4M is an extended version of the 4M multimodal model that combines the current 4M modalities with new video and audio data to provide environment descriptions aimed at helping individuals with impaired vision navigate their surroundings. This project was done as team of four for the course "Communications Project" at EPFL.

Website PDF
Paper thumbnail

Entity Recognition

I experimented with multiple different methods of constructing demonstrations and prompts for the OpenAI LLM to try to improve its performance on token classification. I explored utilizing LLMs themselves as auxiliary tools within the overall prompt engineering process.

PDF

Personal

I collect a list of my favorite CS comic strips here.

These are some of my favorite pictures from the places I've been.

Also, I love Adam ♥