Jane Researcher

Research

Publications, preprints, and ongoing projects spanning machine learning, causal inference, and health informatics.

BenchmarksCausal InferenceDistributed SystemsEHRFederated LearningHealth InformaticsLLMsMachine LearningNLPPrivacyScientific AIStatistics
Large Language Models for Scientific Reasoning: Benchmarks and Limitations
published
Jane Researcher, Carlos Mendez, Aiko Tanaka·Nature Machine Intelligence·

We introduce SciReason-1K, a benchmark of 1,000 expert-curated problems spanning physics, chemistry, and biology, and evaluate state-of-the-art LLMs against domain specialists. Our findings reveal systematic gaps in causal reasoning and quantitative inference that persist across model scales.

NLPLLMsScientific AIBenchmarks
Causal Inference in Observational Health Data: A Practical Framework
published
Jane Researcher, Priya Kapoor, Marcus Webb·Journal of the American Medical Informatics Association·

We propose a unified pipeline for causal effect estimation in large electronic health record datasets, combining propensity score weighting with doubly-robust estimators and sensitivity analyses for unmeasured confounding. Validated on three real-world cohorts.

Causal InferenceHealth InformaticsStatisticsEHR
Federated Learning Under Realistic Non-IID Conditions: A Comprehensive Study
in-review
Jane Researcher, Dmitri Volkov, Soo-Yeon Park·ICML 2025·

We systematically study how label imbalance, feature shift, and temporal drift individually and jointly affect federated learning convergence and fairness. We propose FedAdapt, an adaptive aggregation strategy that reduces worst-silo accuracy gaps by up to 40%.

Federated LearningPrivacyMachine LearningDistributed Systems