Loading...
Loading...
Student performance, MOOC logs, knowledge tracing, standardized tests, learning analytics
13,786 datasets
A text dataset for instruction tuning of language models, created by kshitijthakkar and last updated in December 2025. The dataset contains at least 66,652 examples, with token statistics indicating a focus on long-context training. The total token count is 277,312,737, with a mean of 4,160 tokens per example.
WithinUsAI's Elite God Coder 100K dataset provides 100,000 examples intended for supervised fine-tuning of large language models for advanced software development behavior. The dataset is structured as JSONL files, including a training set, validation set, and a 100-example sample. It was last updated on December 30, 2025.
The boundaries of Dutch municipalities from 1812 onward, including names, Amsterdam Code, and CBS code. It includes the observation year for each polygon to facilitate matching. The data is available in GeoJSON format for use in tools like QGIS, Python, and R.
CitationGround-1M is a citation-locked grounded QA/RAG dataset requiring answers to be based solely on provided contexts. It includes span-level citations and answerable=false hard negatives for training abstention behavior. The dataset was created by WithinUsAI and published on Hugging Face in December 2025.
Giving access to replication data for a neural potential field method used in high-fidelity gravity modeling and density inversion of irregular small celestial bodies. The data supports the research published in Engineering Applications of Artificial Intelligence in 2026, authored by Gao, Wutong.
A 5-year dataset covering admissions and merit at NED University. The dataset is hosted on Kaggle and is intended for analysis of educational selection processes. The specific author, organization, and detailed structure are not provided in the available metadata.
Comprising an oral history interview with Norman Tennent, Professor Emeritus of Conservation Science at the University of Amsterdam. The interview is a retrospective on his education, career, and teaching, exploring ties between personal values and research goals. The specific data structure, including row and column counts, is unknown.
An oral history interview with private conservator Michel van de Laar, conducted by K. Morgan for the University of Amsterdam's Oral History and Artwork Life Stories project. The interview covers his career trajectory and perspectives on the restorer's role. The dataset was last updated on January 27, 2026.
FEDERATED DEEP LEARNING FOR REAL‑TIME DETECTION OF MULTIDRUG‑RESISTANT TUBERCULOSIS is a dataset from Kaggle. It is associated with research into applying federated learning for medical detection tasks. The dataset's specific contents, size, and origin are not detailed in the provided metadata.
A dataset for quantifying the endoplasmic reticulum–mitochondria encounter structure (MAM) using deep learning. The data was uploaded to Kaggle and is tagged for research purposes. The specific source, size, and collection date are not provided in the metadata.
A research dataset from Kaggle proposing a hybrid framework combining physics-based models with deep learning for adaptive control of dielectric materials. The dataset likely contains simulation or experimental data to support the development and validation of the described control framework. Specific details on volume, authorship, and timeliness are not provided in the input.
Deep Learning‑Driven Multiplexed Mass Cytometry data is hosted on Kaggle. The dataset is described as enabling the accurate assessment of biological age from circulating immune cells. The specific volume, author, and update details are not provided.
Data-Driven Predictive Fatigue Assessment for Ship Hull Panels is a dataset from Kaggle intended for research in structural engineering. The description suggests it contains data for modeling fatigue in ship hull components. The specific data volume, source, and creation date are not provided.
A research framework for simultaneous analysis of remote sensing data, developed by J. E. Lang, Ph.D. from the Remote Sensing Systems Laboratory at an unspecified university. The dataset's specific size, temporal coverage, and geographic scope are not detailed in the provided metadata.
Memory-Efficient Gradient Accumulation and Defer is a research dataset from Kaggle. It addresses GPU memory constraints in large-scale deep-learning models. The dataset's author, organization, and specific data content are not detailed in the provided metadata.
Mathematical problems from the 2025 American Invitational Mathematics Examination (AIME), featuring fewer than 1,000 records. It was released in 2025 by Yifan Zhang and the Math-AI team to support research in automated mathematical reasoning.
Data from data.cityofnewyork.us provides counts of students participating in high school admissions, taking the SHSAT, and receiving offers for each public NYC middle school. The dataset includes columns for Feeder School Name, Feeder School DBN, Year, Count of Students in HS Admissions, Count of Testers, Number of Offers, and Number of Discovery participants. It was last updated on 2025-12-04 21:49:03.
An educational guide for newborn resuscitation, developed for midwifery students. It was created through workshops involving midwives and students using the ADDIE instructional design model. The dataset's specific row count, column count, and file structure are unknown.
25,000 examples designed to train and evaluate master-scholar reasoning about the Universe. The dataset spans rigorous physics, observational astronomy, cosmology, and methodological/epistemic foundations. It was created by WithinUsAI and last updated on January 3, 2026.
Encompassing primary research conducted in March 2025 on the relationship between parental and community involvement and learner discipline in secondary schools in Kiruhura District, Uganda. The author is Gidraf Joseph Wanjala, and the data was last updated in February 2026. Specific details on the number of rows, columns, and file formats are unavailable.