Round 4 Train DatasetThe data being generated and disseminated is the training data used to construct trojan detection software solutions. This data, generated at NIST, consists of human level AIs trained to perform image classification. 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 3 Training DatasetThe data being generated and disseminated is the training data used to construct trojan detection software solutions. This data, generated at NIST, consists of human level AIs trained to perform image classification. 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 2 Test DatasetThe data being generated and disseminated is the test data used to evaluate trojan detection software solutions. This data, generated at NIST, consists of human level AIs trained to perform a variety of tasks (image classification, natural language processing, etc.). 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 2 Holdout DatasetThe data being generated and disseminated is the holdout data used to evaluate trojan detection software solutions. This data, generated at NIST, consists of human level AIs trained to perform a variety of tasks (image classification, natural language processing, etc.). 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 3 Test DatasetThe data being generated and disseminated is the training data used to construct trojan detection software solutions. This data, generated at NIST, consists of human level AIs trained to perform image classification. 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 3 Holdout DatasetThe data being generated and disseminated is the training data used to construct trojan detection software solutions. This data, generated at NIST, consists of human level AIs trained to perform image classification. 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.

The REsource for Materials Informatics (REMI) will host a diverse collection of scripting notebooks (Jupyter, Matlab LiveScripts, etc.) for collecting, pre-processing, analyzing, and visualizing materials data. Notebooks are curated using tags aligned to Materials Science and Data Science topics. REMI emerged from the realization that both experts and novices wanted examples of using machine learning for science. Meanwhile, lots of experts are developing digital notebooks (e.g. Jupyter) to demonstrate step-by-step data collection, pre-processing, analysis and visualization.

Code and demonstration data for the paper, "On-the-fly closed-loop materials discovery via Bayesian active learning," Kusne, A.G., Yu, H., Wu, C. et al. Nat Commun 11, 5966 (2020). https://doi.org/10.1038/s41467-020 19597-w Code: Closed-loop autonomous materials exploration and optimization. This code is used to control an autonomous materials exploration and optimization platform. It guides subsequent experiments to learn about a material's phase map and target functional properties in a unified framework.

This data set consists of both measured and simulated optical intensities scattered off periodic line arrays, with simulations based upon an average geometric model for these lines. These data were generated in order to determine the average feature sizes based on optical scattering, which is an inverse problem for which solutions to the forward problem are calculated using electromagnetic simulations after a parameterization of the feature geometry.

Round 2 Training DatasetThe data being generated and disseminated is the training data used to construct trojan detection software solutions. This data, generated at NIST, consists of human level AIs trained to perform image classification. 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.