CV

• Developed a theory of mind benchmark measuring model reasoning as a graded construct: 201 items across 12 dimensions, each mapped to empirically established human age norms (ages 2.5 to 11) to enable developmental age-equivalent scoring of large language models.
• Built evaluation harness in Inspect (AISI), running 28 models across 5 providers from a 53-model routing panel (5,000+ graded observations) with custom datasets, solvers, and scorers, reusable across providers without per-model rewrites. Handled multi-provider model configuration, rate-limit and retry behavior, and stateless calls under zero-data-retention.
• Red-teamed items against contamination and shortcut solutions, rewriting stems that leaked answers through lexical overlap and constructing distractors plausible under the wrong reasoning path.
• Ran dual analytic pipelines over the corpus, combining mastery-based age mapping with mixed-effects and IRT model estimation including task-format covariates, so that scores are not confounded with prompt format.
• Found that no model tier in any of 5 families follows the human acquisition sequence: profiles are non-monotonic across all 12 dimensions, with sizeable age-equivalent-year swings between adjacent skills, evidence that models reach competencies while bypassing their logical prerequisites.
• Created an interactive Streamlit dashboard served by a Parquet snapshot, making results explorable beyond the write-up.

• Built the import and validation layer for a multi-lab open repository standardizing eye-tracking data across 12+ labs in a shared open-access database, enabling cross-study comparisons.
• Defined relational schema, wrote extract-transform-load scripts for 5 contributed datasets, and added automated schema checks flagging protocol mismatches at ingestion.

• Jointly embedded mobile phone and eye-tracking data in a low-dimensional manifold space to classify attentional engagement; demonstrated improved classification accuracy over baseline feature-space methods.

• Built and deployed a full-stack web and iOS application enabling family groups to set, track, and share daily and weekly health goals; integrated with the iOS Health API for automated step and activity syncing; includes password-protected group accounts and custom goal assignment.

• Building a browser-based implementation of the validated ToM Booklet task (Sotomayor-Enriquez et al., 2023), with audio narration, animated stimuli, and automated response capture to enable large-scale online data collection.

• Doctoral-track Candidate in Psychology (09/2023–10/2025)
• Project: Structural and Functional Neural Mechanisms of Theory of Mind Development
• M.Sc. in Psychological Research, Awarded With Merit (11/2023)
• Dissertation: Studying the distinct functions of theory of mind brain regions during movie-viewing
• UKRI ESRC DTP Studentship (09/2022–10/2025)

• B.A. in Computer Science (Harvey Mudd College) and Neuroscience (Keck Science), Awarded Cum Laude (05/2021)
• Thesis: Uncovering Object Categories in Infant Views
• NSF REU Scholarships: Stanford (CSLI), 02/2020; University of Minnesota (Neuroimaging), 05/2019
• Scripps College Humanities Institute Fellowship (01/2021)
• Scripps College Success Grant Scholarship (08/2020, 08/2021)

• Designing evaluation rubrics for frontier language models informed by developmental cognition research, targeting reasoning and coding tasks where model failure modes parallel challenges in human cognitive development.
• Improved rubric reliability and evaluator agreement through structured feedback on task specifications, isolating cases where scale ambiguity rather than response quality drives annotation variance.

• Built reproducible Python pipelines for functional and diffusion MRI data (fMRIPrep, FSL, MRtrix3) to study the structural and functional neural mechanisms of theory of mind development in 60 children aged 5 to 12.
• Led workshops and hackathons to align researchers, clinicians, and technical contributors across concurrent projects, teaching data processing, organization, and analysis for studying cognitive development.

• Designed and scaled a code review program for researchers across 3 departments, establishing shared standards for publication-ready projects across contributors with varying engineering experience.
• Advised researchers on open science practices at all stages of the research lifecycle, hosting workshops and 1:1 consultations on data governance, preregistration, and reproducible research design.

• Owned collection, processing, and analysis workflows for multimodal data, including online experiments, eye-tracking, and functional MRI.
• Collected data for and contributed to the write-up of a study showing how novelty and uncertainty differentially drive information-seeking exploration across development.
• Managed IRB/ethics compliance, budgeting, and codebase for research on adolescent decision-making, reinforcement learning, and mental health risk factors.
• Led school outreach and participant recruitment with NYC public and private schools, coordinating with educators and caregivers to support safe, informed participation of minors.

• Re-engineered the participant database system in Python for a lab managing 500+ active participants, replacing manual tracking with queryable infrastructure and reducing coordination overhead.

• Fine-tuned and tested Detectron2 object detection and segmentation models in PyTorch on 10,000 annotated frames of infant egocentric video to characterize early visual experience and evaluate model performance on unusual image angles.

• Built human-in-the-loop annotation workflows with AWS SageMaker to label egocentric videos from infants' home environments, producing a training set for object recognition models.

• Trained supervised classifiers to predict perceived color using functional MRI data.

Nussenbaum, K., Martin, R. E., Maulhardt, S., Yang, Y., Bizzell-Hatcher, G., Bhatt, N. S., Scheuplein, M., Rosenbaum, G. M., O'Doherty, J. P., Cockburn, J., & Hartley, C. A. (2023). Novelty and uncertainty differentially drive exploration across development. eLife.

Zettersten, M., …, Bhatt, N. S., Bergey, C. A., & Frank, M. C. (2022). Peekbank: An open, large-scale repository for developmental eye-tracking data of children's word recognition. Behavior Research Methods.

Long, B., Kachergis, G., Bhatt, N. S., & Frank, M. C. (2021). Characterizing the object categories two children see and interact with in a dense dataset of naturalistic visual experience. Proceedings of the 43rd Annual Conference of the Cognitive Science Society.

• Designed and delivered two interdisciplinary lectures bridging expertise in cognitive neuroscience and AI for advanced secondary-school students.
• Brain-Inspired Artificial Intelligence (slides): led discussion-based lesson on the neurobiological visual system (covering receptive fields, Hubel & Wiesel, ventral pathway), and mapped those mechanisms to the history of neural networks (i.e., McCulloch–Pitt threshold units, Hebbian learning, Rosenblatt/multi-layer perceptrons, and convolution).
• Applications of AI: Cognitive Science (slides): taught interdisciplinary foundations of cognitive science (e.g., brain versus mind), introduced functional MRI methodologies, and connected neuroimaging data to machine learning tools, such as multivoxel pattern analysis and support vector machines.

• Data Analysis for Psychology in R 1 (PSYL08013) with Dr. Umberto Noe, Autumn 2023 – Spring 2025: mentored students through group-based pair programming assignments, teaching fundamentals of univariate statistics and R programming.
• Psychology 1A/B with Dr. Hannah Cornish, Autumn 2023 – Spring 2024.
• Data Analysis for Psychology in R 3 (PSYL10168) with Dr. Umberto Noe, Autumn 2023: mentored students through fundamental questions relating to multivariate statistics and R programming.

• Computability and Logic (CSCI-081) with Dr. George Montañez: led weekly sessions and tutored 20 students to clarify topics relating to computability theory, set countability, formal logics (predicate, propositional, etc.), Chomsky grammars and languages, Turing machines, and automata; continued teaching virtually during pandemic.

• Calculus II w/ Applications to Science (MATH-031S) with Dr. Blerta Shtylla: led twice-weekly sessions and tutored 30 students to help understand series and sequences, integration, differential equations, and probabilistic applications using examples in the sciences.

• Elected by 30+ postgraduate researchers to represent student interests in faculty board meetings and university-level committees; successfully advocated to preserve community funding amidst university-wide budget cuts.

• Designed and presented an invited workshop on computational modeling for cognitive neuroscience for attendees at an international conference.

• Founded and led a program serving underrepresented high school students across Los Angeles and the Inland Empire.

• Supported community-building and logistics for workshop attendees within the broader NeurIPS research community.

• Human Research: Social & Behavioral Research Investigators (Refresher Course), Massachusetts Institute of Technology, 10/2022; expired 10/2025 · Certificate (PDF) ↗ · Verify ↗
• Revised Common Rule: Investigators (Basic Course), New York University, 05/2021; no expiration · Certificate (PDF) ↗ · Verify ↗
• Human Research: Social/Behavioral or Humanist Research Investigators and Key Personnel (Basic Course), University of Minnesota, 06/2019; expired 06/2022 · Certificate (PDF) ↗ · Verify ↗
• Good Clinical Practice and Human Research Protections for Biomedical Study Teams (Basic Course), University of Minnesota, 06/2019; expired 06/2022 · Certificate (PDF) ↗ · Verify ↗
• Responsible Conduct of Research: Research Involving Human Subjects (Basic Course), University of Minnesota, 06/2019; no expiration · Certificate (PDF) ↗ · Verify ↗
• Responsible Conduct of Research: Agency Specific (Basic Course), University of Minnesota, 06/2019; no expiration · Certificate (PDF) ↗ · Verify ↗