No abstract

The database of Global Ionospheric Maps (GIMs) produced at Jet Propulsion Laboratory is analyzed. We define high density Total Electron Content (TEC) regions (HDRs) in a map, following certain selection criteria. For the first time, we trained four convolutional neural networks (CNNs) corresponding to four phases of a solar cycle to classify the GIMs by the number of HDRs in each map with ∼76% accuracy on average. We compared HDR counts for GIMs across ten years to draw conclusions on how the number of HDRs in the GIMs changes throughout the solar cycle. Occurrence of HDRs during different geomagnetic activity conditions is discussed. Catalogue of selected HDRs for ten years and four CNN-based models that can be used to extend classification to other years are provided for the community to use.

Prompting interfaces allow users to quickly adjust the output of generative models in both vision and language. However, small changes and design choices in the prompt can lead to significant differences in the output. In this work, we develop a black-box framework for generating adversarial prompts for unstructured image and text generation. These prompts, which can be standalone or prepended to benign prompts, induce specific behaviors into the generative process, such as generating images of a particular object or biasing the frequency of specific letters in the generated text. * Equal contribution.

No abstract

No abstract