Round 10 Train DatasetThis is the training data used to create and evaluate trojan detection software solutions. This data, generated at NIST, consists of object detection AIs trained on the COCO dataset. A known percentage of these trained AI models have been poisoned with a known trigger which induces incorrect behavior. This data will be used to develop software solutions for detecting which trained AI models have been poisoned via embedded triggers. This dataset consists of 144 AI models using a small set of model architectures.

Round 9 Test DatasetThis is the test data used to evaluate trojan detection software solutions. This data, generated at NIST, consists of natural language processing (NLP) AIs trained to perform one of three tasks, sentiment classification, named entity recognition, or extractive question answering on English text. A known percentage of these trained AI models have been poisoned with a known trigger which induces incorrect behavior. This data will be used to develop software solutions for detecting which trained AI models have been poisoned via embedded triggers.

Round 9 Holdout DatasetThis is the holdout data used to evaluate trojan detection software solutions. This data, generated at NIST, consists of natural language processing (NLP) AIs trained to perform one of three tasks, sentiment classification, named entity recognition, or extractive question answering on English text. A known percentage of these trained AI models have been poisoned with a known trigger which induces incorrect behavior. This data will be used to develop software solutions for detecting which trained AI models have been poisoned via embedded triggers.

Speed of Sound Measurements of Binary Mixtures of Difluoromethane (R-32) with 2,3,3,3-Tetrafluoropropene (R-1234yf) or trans-1,3,3,3-Tetrafluoropropene (R-1234ze(E)) Refrigerants

Round 8 Train DatasetThis is the training data used to construct and evaluate trojan detection software solutions. This data, generated at NIST, consists of natural language processing (NLP) AIs trained to perform extractive question answering (QA on English text. A known percentage of these trained AI models have been poisoned with a known trigger which induces incorrect behavior. This data will be used to develop software solutions for detecting which trained AI models have been poisoned via embedded triggers.

Here we provide curated analytical chemistry data for eggs collected from 1999 to 2010 on a subset of species and analytes that were measured regularly and reasonably systematically. Included in this publication are 487 samples analyzed for 174 ubiquitous environmental contaminants such as (poly)brominated diphenyl ethers (BDEs), mercury, organochlorine pesticides (OCPs), and polychlorinated biphenyls (PCBs). Data were collated to form a dataset useful in chemometric and related analyses of the marine ecosystem in the north Pacific Ocean.

Data from peer-reviewed publication: G. Zhao et al., Frequency stabilization of a quantum cascade laser by weak resonant feedback from a Fabry-Perot cavity, Optics Letters. Frequency-stabilized mid-infrared lasers are valuable tools for precision molecular spectroscopy. However, their implementation remains limited by complicated stabilization schemes. Here we achieve optical self-locking of a quantum cascade laser to the resonant leak-out field of a highly mode-matched two-mirror cavity.

Data corresponding to Figs. 2,3,5,6,11,12,13,14 for Fleurbaey et al J. Quant Spectrosc. Radiat. Transf. vol 270, 107684 (2021). doi.org.10.1016/j.jqsrt.2021.107684

Dataset for generation of figures in Air-broadening in near-infrared carbon dioxide line shapes: quantifying contributions from O2, N2, and Ar (https://doi.org/10.1016/j.jqsrt.2021.107669)

Round 7 Holdout DatasetThis is the holdout data used to construct and evaluate trojan detection software solutions. This data, generated at NIST, consists of natural language processing (NLP) AIs trained to perform named entity recognition (NER) on English text. A known percentage of these trained AI models have been poisoned with a known trigger which induces incorrect behavior. This data will be used to develop software solutions for detecting which trained AI models have been poisoned via embedded triggers.