Research Track 01
On the Robustness of LLMs' Internal Representation of Code Correctness
Francisco Ribeiro, Sohaila Abdulsattar, Renata Gonzalez, Mahmoud Kassem, Sarah Nadi
I'm a Global PhD Fellow in Computer Science at NYU. I work on AI for software engineering, with a focus on improving LLMs and agents for code generation.
Improving how language models generate code and understanding where they fail, with the goal of making the code they produce more reliable
Evaluating how LLM-based agents handle larger, multi-step software engineering tasks, and what limits them in real projects
Interpreting the internal representations of language models to understand what they encode about code and how that relates to their behavior
Using reinforcement learning to strengthen the reasoning abilities of language models, and studying how reward signals shape the code they produce
Francisco Ribeiro, Sohaila Abdulsattar, Renata Gonzalez, Mahmoud Kassem, Sarah Nadi
Sohaila Abdulsattar, Keith Ross
New work is on the way. If you'd like to collaborate or talk research, I'd love to hear from you!
"On the Robustness of LLMs' Internal Representation of Code Correctness" will appear in the ICSE 2027 Research Track.
A student organization that builds community and support for underrepresented groups in NYU's CS PhD program.
Joined the SANAD Lab, starting the program in New York.
Bachelor of Science in Computer Science, with a minor in Applied Mathematics.
"Arabic Dialect Translation with Small LLMs: Enhancing through Reasoning-Oriented Reinforcement Learning," published in the EACL 2026 workshop proceedings.
I've always had an eye for design, and I'll take any excuse to make something look elegant. It's also why this website turned out pretty cool, if I do say so myself ;)
Badminton is the one sport I've actually gotten better at. I'm still not especially great, but I take it as seriously as someone who is.
I've loved Minecraft since I was a kid and somehow never grew out of it. These days I watch more than I play, and I love finding out about new mods.