Experience

Postdoc

New York University Abu Dhabi (NYUAD)
July 2024 – Present

Conducting research on LLMs for code generation and improving reliability of LLM-generated code through representation engineering and interpretability techniques.

Research Internship

National Institute of Informatics, Tokyo, Japan

• January 2024 – March 2024: Developing GPT-4-based assistants for legal debugging
• September 2022 – January 2023: Using GPT-3 and type systems to automatically locate and repair type errors

Teaching / Supervising

University of Minho, Braga, Portugal (2019 – 2023)

• Teaching Assistant for 1st and 2nd year courses for which I had to cover all 3 main programming paradigms:
  • Functional
  • Imperative
  • Objected-Oriented
• Co-supervision of MSc dissertation
  • Simão Cunha: Towards a Platform for Benchmarking Large Language Models (co-supervised with João Saraiva; 2023 — 2024)