About Me

I am a second year PhD student advised by Arjun Guha at the Programming Research Laboratory of Northeastern University. I am interested in the intersection of NLP, programming languages and systems. I work on aligning CodeLLMs with the goal of improving robustness and decreasing model size for resource constrained applications. Most recently, I worked on semi-synthetic data augmentation for improving the performance of CodeLLMs on low-resource programming languages. My research is based on mechanistic interpretability, or the view that capabilites of transformers arise from specific model components.

Publications

• Lucchetti, F., & Guha, A. (2024). Activation Steering for Robust Type Prediction in CodeLLMs. ArXiv preprint.

• Hu, Z., Lucchetti, F., Schlesinger, C., Saxena, Y., Freeman, A., Modak, S., … & Biswas, J. (2024). Deploying and Evaluating LLMs to Program Service Mobile Robots. IEEE Robotics and Automation Letters.

• Cassano, F., Gouwar, J., Lucchetti, F., Schlesinger, C., Anderson, C. J., Greenberg, M., … & Guha, A. (2023). Knowledge Transfer from High-Resource to Low-Resource Programming Languages for Code LLMs. Arxiv preprint.

• Lucchetti, F., Hu, Z., Freeman, A., Modak, S., Saxena, Y., Mao, L., Schlesinger, C., Guha, A., Biswas, J. (2023). Deploying and Evaluating LLMs to Program Service Mobile Robots. IROS workshop.