Associate Professor of Computer Science
University of Colorado Boulder
I study the production and evaluation of scientific knowledge using machine learning and computational social science.
arXiv preprint arXiv:2605.29468 · 2026 · Preprint
Do language models uphold research-integrity norms when misconduct is framed as a routine request?
Across 16 models, refusals were substantially more reliable for explicit misconduct than for covert violations. SciIntBench evaluates both misconduct refusal and helpfulness on legitimate scientific tasks.
arXiv preprint arXiv:2604.00248 · 2026 · Preprint
Can training rewards help generated reviews use figures and external scholarly context?
An 8B-parameter model trained with complementary correspondence rewards achieved the highest review-quality scores among the six baselines evaluated in the study.
Science Advances · 2025
Media: ScienceNatureThe ScientistChemistry WorldPhysics WorldTechnology NetworksThe RegisterDigital JournalCU Boulder TodayScienceDailyNeuroscience NewsCan journal websites and publication metadata help identify journals that warrant closer investigation?
At a 50% decision threshold, the classifier flagged 1,437 of 15,191 journals. Manual review estimated that 24% of flagged journals were false positives, supporting automated screening as a guide to expert investigation.
My research examines peer review, research integrity, scientific careers, and collaboration. I use scholarly data, statistical models, and machine learning to study these processes.
I lead the Science of Science and Computational Discovery Lab at CU Boulder. Our work includes empirical studies of scientific institutions and computational methods for analyzing scientific text, images, and citation networks.
Research overviewMethods for detecting figure reuse, characterizing questionable journals, and evaluating the behavior of AI systems in research settings.
Related publications Estimating the predictability of questionable open-access journalsScience Advances · 2025Studying how funding, mentorship, collaboration, and recognition shape scientific work—and who gets the opportunity to contribute.
Read the research Dissecting the gender divide: authorship and acknowledgment in scientific publicationsScientometrics · 2026Building systems that help researchers find relevant work, identify unanswered questions, and evaluate scientific manuscripts.
Read the research REM-CTX: Automated Peer Review via Reinforcement Learning with Auxiliary ContextarXiv · 2026 · PreprintUniversity of Colorado Boulder
Three-course specialization · Supervised learning, unsupervised learning, and deep learning
Before joining CU Boulder, I was a faculty member in Syracuse University’s School of Information Studies. I earned my Ph.D. in Computer Science at the University of Minnesota and was a postdoctoral researcher at Northwestern University and the Rehabilitation Institute of Chicago.
I was born in Santiago, Chile, and studied at the University of Santiago. I also contribute to the Science of Science Summer School and the Computational Research Integrity Conference.