Loading...
Loading...
Text classification, translation, QA, summarization, dialogue, sentiment analysis, language modeling, text corpora
46,416 datasets
A discussion paper from the Australian Ocean Data Network presents geological fault and fold data for Papua New Guinea. The data was compiled from ground surveys, remote sensing, and oil exploration company work over the last 25 years, culminating in a 1:1,000,000 scale geological map. It specifically critiques and corrects a 1971 publication by Wellman regarding lineaments and fault systems in the region.
ExtremeOcc-3D is a benchmark introduced with a 2026 ICML Spotlight paper. It contains 264 single-view indoor scenes for evaluating amodal 3D generation under severe object occlusion. Each sample provides rendered images, masks, text labels, and ground-truth 3D assets.
NOR Units data from the Texas Commission on Environmental Quality details waste management unit registrations across the state. It includes flags for hazardous and industrial waste classes, unit descriptions, capacity, and regulatory statuses. The dataset is maintained by the state of Texas and was last updated in April 2026.
A study of sedimentology and Holocene history for the Broad Sound tropical estuary in Queensland, Australia. The dataset likely contains information on sediment types, grain size parameters, and radiocarbon dating from drilling. It was published by the Australian Ocean Data Network and last updated in April 2026.
A geological report from the Australian Ocean Data Network details the stratigraphy and structure of the Ngalia Basin in Australia's Northern Territory. The document describes up to about 5 km of sedimentary rocks, including thicknesses for Adelaidean (3200 m), Cambrian (800 m), probable Ordovician (300 m), and Devono-Carboniferous (3100 m) sequences. It was last updated on 2026-04-10.
A newly compiled structural elements map and chronostratigraphic section illustrate the basin's complex development through at least two major extensional faulting periods. The Australian Ocean Data Network provides this analysis, which includes palaeogeographic maps in a plate reconstructed setting. The dataset was last updated on April 10, 2026.
Tianhao Tan published a dataset on figshare in April 2026 for revealing design rules for bathochromic narrowband emission in ICz-based MR-TADF emitters. The dataset, 120.1 KB in size, contains calculated absorption and emission spectra. Calculations were performed using the TVCF method in MOMAP software, with electronic energy and structure computed at the B3LYP-D3(BJ)/6-31G* level.
Australian Ocean Data Network provides a geological report on Phanerozoic rocks in onshore Western Australia. The description details the sedimentary, volcanic, and intrusive rock distributions across major basins like the Bonaparte, Carnarvon, Canning, and Perth Basins, linking their formation to the separation from Gondwana. It also catalogues associated mineralisation, including base metals, coal, mineral sands, evaporites, diamonds, and iron ore.
A reconnaissance study of the major Palaeozoic aquifers in the Canning Basin, Western Australia, based on data from 30 oil exploration wells. The 430,000 km2 basin contains Zn-Pb sulphide deposits, and the report analyzes groundwater flow, temperature, and salinity to model ore-forming fluid mechanisms. The dataset was published by the Australian Ocean Data Network and last updated in April 2026.
PREFIRE Satellite 2 Raw Payload contains uncurated digital number counts from a Thermal Infrared Spectrometer (TIRS-PREFIRE) with 63 channels measuring radiation from 5 to 53 µm. The data aims to fill knowledge gaps in the polar far-infrared energy budget for assimilation into global circulation and climate models. Measurements began on June 29, 2024, have a temporal resolution of 0.707 seconds, and provide global coverage with a concentration in polar regions.
Global satellite data, with a concentration over polar regions, provides raw, uncurated digital number counts from a 63-channel push broom spectrometer measuring mid- to far-infrared radiation (5-53 µm). This data, collected at a 0.707-second temporal resolution starting in July 2024, aims to fill knowledge gaps in the polar radiant energy budget. The measurements are intended for assimilation into global circulation and climate models to improve future climate predictions.
A collection of query texts from the full works of Hieronymus (Jerome) and Lactantius, used for finding intertextual relationships with classical Latin literature. The dataset is part of a larger collection focused on detecting intertextual connections between late-antique citing authors and classical Latin literature. It was created by julian-schelb and was last updated on 2026-05-16.
A 25,000-example instruction-tuning dataset designed to train language models to embody the persona of Archangel Metatron. The dataset was created by author gss1147 and was last updated on Hugging Face in May 2026. Its purpose is to instill the perspective, voice, and energetic signature associated with this spiritual figure.
50 document samples were processed on 2026-05-21 using the NuExtract3 vision-language model for structured extraction. The output dataset is derived from the 'davanstrien/doab-nuextract3-smoke-input' source and was created by the user davanstrien. Processing was completed in 5.3 minutes using the 'structured-extraction' mode.
47.0 KB of experimental data for a functionally heterogeneous Prussian Blue-based nanozyme array. The dataset, authored by yang guan chen and last updated in April 2026, is structured across seven Excel sheets for machine learning analysis and figure generation. It likely contains measurements for qualitative screening and semi-quantification of polymetallic mixtures.
57 programming languages are represented in The Heap, a dataset designed for reproducible evaluation of large language models. It was developed by AISE-TUDelft and last updated in May 2026. The dataset is described as contamination-free, aiming to prevent data leakage between training and evaluation sets.
71 Bryoria lichen specimens from North America, Europe, and Asia were sequenced for six gene regions to examine infrageneric relationships. The study by Myllys, Leena reveals a recent diversification in section Divaricatae, estimated at around 5 million years old, and describes a new species, B. ahtiana. Data includes alignments and phylogenetic analyses from seven datasets prepared for maximum parsimony, likelihood, and Bayesian methods.
Mathew Thomas created a dataset integrating LiDAR-derived shade frequency, Landsat land surface temperature, and 2021 Statistics Canada census indicators. It constructs three indices—Exposure, Social Equity, and Adaptive Capacity—and a Composite Vulnerability Index (CVI) for six University of British Columbia Vancouver neighbourhoods. The dataset was last updated on May 2, 2026, and identifies East Campus and University Boulevard as the most vulnerable areas.
May to July 2022 measurements of methane screening (ppm) and methane emission rates (mg/h) collected from non-producing oil and gas wells in Alberta and Saskatchewan. The dataset also includes average daily wind speed and maximum wind gust speed. It was authored by Bianca Lamarche and hosted on the Borealis Harvested Dataverse.
Fourteen years of monthly MODIS sea surface temperature data were used to map coastal upwelling frequency. The GIS layer distinguishes areas of persistent, semi-persistent, and seasonal upwelling along the south-eastern Australian coast. The data was generated by researchers Huang and Wang and is hosted by the Australian Ocean Data Network.