TrojAI llm-pretrain-apr2024 Train DatasetThis is the training data used to create and evaluate trojan detection software solutions. This data, generated at NIST, consists Llama2 Large Language Models refined using fine-tuning and LoRA to perform next token prediction. A known percentage of these trained AI models have been poisoned with triggers which induces modified behavior. This data will be used to develop software solutions for detecting which trained AI models have been poisoned via embedded triggers into the model weights.

TrojAI cyber-network-c2-mar2024 Train DatasetThis is the training data used to create and evaluate trojan detection software solutions. This data, generated at NIST, consists of ResNet18 and ResNet34 neural network models that classify botnet command and control (c2) and benign network traffic packets trained on the USTC-TFC2016 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.

Round rl-randomized-lavaworld-aug2023-train Train DatasetThis is the training data used to create and evaluate trojan detection software solutions. This data, generated at NIST, consists of Reinforcement Learning agents trained to navigate the Lavaworld Minigrid environment. 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.

nlp-question-answering-aug2023-trainThis is the train data used to evaluate trojan detection software solutions. This data, generated at NIST, consists of natural language processing (NLP) AIs trained to perform 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 rl-lavaworld-jul2023-train Train DatasetThis is the training data used to create and evaluate trojan detection software solutions. This data, generated at NIST, consists of Reinforcement Learning agents trained to navigate the Lavaworld Minigrid environment. 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.

TrojAI cyber-apk-nov2023 Train DatasetThis is the training data used to create and evaluate trojan detection software solutions. This data, generated at NIST, consists of small feed forward multi-layer perceptron type neural network models classifying APK feature vectors as malware or clean. 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.

TrojAI cyber-network-c2-feb2024 Train DatasetThis is the training data used to create and evaluate trojan detection software solutions. This data, generated at NIST, consists of ResNet18 and ResNet34 neural network models that classify botnet command and control (c2) and benign network traffic packets trained on the USTC-TFC2016 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.

Round 13 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 both on synthetic image data build from Cityscapes and the DOTA_v2 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 128 AI models using a small set of model architectures.

Data for "Preparing for 6G: Developing best practices and standards for industrial measurements of low-loss dielectrics"

Round 12 Train DatasetThis is the training data used to create and evaluate trojan detection software solutions. This data, generated at NIST, consists of pdf malware classification AIs trained Contaigio dataset feature vectors. 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 120 AI models using a small set of model architectures.