Cedrick Argueta
Graduate Researcher
New Jersey, United States
Summary
👤
Senior🎓
Top SchoolCedrick Argueta is a graduate researcher at Princeton with nine years of hands-on experience applying deep reinforcement learning to robotics, safety, and adversarial ML. Trained at Stanford (CS, AI concentration) and now pursuing a PhD, he has industrial experience at The Aerospace Corporation and NASA JPL building RL systems and automated test infrastructure for aerospace platforms. He contributes to prominent open-source evaluation work—helping implement the "convinceme" task in Google's BIG-bench—demonstrating expertise in Python, ML evaluation, and model persuasion metrics. Comfortable spanning research and production code, he blends rigorous academic methods with practical system-building for safety-critical domains. Colleagues describe him as a researcher-engineer who seeks measurable improvements in model behavior, not just novel algorithms.
9 years of coding experience
3 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Princeton UniversityHigh School Diploma, High School Diploma at Abraham Lincoln Senior High SchoolBachelor's degree, Computer Science, concentration in Artificial Intelligence, Bachelor's degree, Computer Science, concentration in Artificial Intelligence at Stanford University
English, Spanish
Github Skills (9)
machine-learning10
python10
apachebench10
json8
nlp7
natural-language-processing7
data-analysis6
scikit-learn6
scikit6
Programming languages (2)
Github contributions (5)
google/BIG-bench
Sep 2021 - Nov 2021
Beyond the Imitation Game collaborative benchmark for measuring and extrapolating the capabilities of language models
Role in this project:
Back-end Developer & Data Scientist
Back-end Developer & Data ScientistContributions:9 reviews, 113 commits, 12 comments in 1 month
Contributions summary:Cedrick primarily contributed to the development of the `convinceme` benchmark task within the `big-bench` repository. Their work involved implementing the core logic of the task, including the integration of a questionnaire to assess jury model beliefs and the calculation of a persuasion score. They leveraged and modified existing self-play task templates and incorporated JSON data for statements. The user demonstrated proficiency in Python and the application of machine learning concepts for evaluating the persuasiveness of models.
bertmachine-learningbenchmarkmeasuringbenchmarks
cdrckrgt/cs221-project
Oct 2018 - Dec 2018
Contributions:43 commits, 28 pushes in 1 month
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.