arXiv subjects

Sophia Abraham

Publications and source records attributed to Sophia Abraham.

6 recordsLinked to original sources

arXiv2026

Silent Updates: Measuring and Closing the Post-Deployment Disclosure Gap

Deployed foundation models are often not static systems, with providers able to modify system behavior through fine-tuning, classifier updates, system prompt revisions, retrieval changes, and routing changes. These updates can be made silently -- that is, without public disclosure, a version increment, or re-evaluation. Such silent updates challenge a core assumption behind current AI governance frameworks that an externally verifiable chain of custody links the model referred to in evaluation results or a system card to the model served to users. In this paper, we examine post-deployment disclosure practices across first-party API providers and inference hosts to establish the extent to which a chain of custody exists in practice. We find that providers commonly publish substantial safety documentation, including quantitative evaluations and version-specific reports, but no provider in our sample published information allowing an external party to verify that the artifact being served is the same one referred to in this documentation. We introduce the Silent Updates Scorecard, a public instrument for measuring post-deployment disclosure practices across providers and hosts, and preliminary results for a sample of nine first-party API providers and seven third-party inference hosts. We also propose a Three-Part Behavioral Trigger System for determining when post-deployment modifications to a system motivate disclosure or re-evaluation obligations.

cs.CY↗

arXiv2026

AI Incident Monitoring through a Public Health Lens

Artificial intelligence systems are now deployed at scale across sectors, accompanied by a growing number of real-world incidents ranging from misinformation and cybercrime to autonomous-system failures. Databases of AI incidents index these events, but they cannot measure ``risk'' (i.e., a joint measure of likelihood and severity) without additional data regarding the prevalence of risk-associated systems and their incident reporting rates. As a result, policymakers, companies, and the general public lack a means to weigh the benefits of AI against their in-context risks. Inspired by public-health processes, which presume noisy and incomplete disease surveillance, we identify six phases of incident emergence. We demonstrate the framework through a detailed case study of autonomous vehicles, whose mandatory reporting requirements produces reliable incident-rate ground truth expressed in distance traveled. The case study shows that an informed panel of domain experts (e.g., self-driving experts) can combine their domain expertise, incident data, and a collection of statistical and visualization tools to arrive at incident phase determinations serving public needs. We further demonstrate the approach with a deepfake incident case study and chart a path for future research in incident phase determination.

cs.CY↗

arXiv2023

Temporal Egonet Subgraph Transitions

How do we summarize dynamic behavioral interactions? We introduce a possible node-embedding-based solution to this question: temporal egonet subgraph transitions.

cs.SI↗

arXiv2021

Adaptive Autonomy in Human-on-the-Loop Vision-Based Robotics Systems

Computer vision approaches are widely used by autonomous robotic systems to sense the world around them and to guide their decision making as they perform diverse tasks such as collision avoidance, search and rescue, and object manipulation. High accuracy is critical, particularly for Human-on-the-loop (HoTL) systems where decisions are made autonomously by the system, and humans play only a supervisory role. Failures of the vision model can lead to erroneous decisions with potentially life or death consequences. In this paper, we propose a solution based upon adaptive autonomy levels, whereby the system detects loss of reliability of these models and responds by temporarily lowering its own autonomy levels and increasing engagement of the human in the decision-making process. Our solution is applicable for vision-based tasks in which humans have time to react and provide guidance. When implemented, our approach would estimate the reliability of the vision task by considering uncertainty in its model, and by performing covariate analysis to determine when the current operating environment is ill-matched to the model's training data. We provide examples from DroneResponse, in which small Unmanned Aerial Systems are deployed for Emergency Response missions, and show how the vision model's reliability would be used in addition to confidence scores to drive and specify the behavior and adaptation of the system's autonomy. This workshop paper outlines our proposed approach and describes open challenges at the intersection of Computer Vision and Software Engineering for the safe and reliable deployment of vision models in the decision making of autonomous systems.

cs.SE↗

arXiv2020

The Next Generation of Human-Drone Partnerships: Co-Designing an Emergency Response System

The use of semi-autonomous Unmanned Aerial Vehicles (UAV) to support emergency response scenarios, such as fire surveillance and search and rescue, offers the potential for huge societal benefits. However, designing an effective solution in this complex domain represents a "wicked design" problem, requiring a careful balance between trade-offs associated with drone autonomy versus human control, mission functionality versus safety, and the diverse needs of different stakeholders. This paper focuses on designing for situational awareness (SA) using a scenario-driven, participatory design process. We developed SA cards describing six common design-problems, known as SA demons, and three new demons of importance to our domain. We then used these SA cards to equip domain experts with SA knowledge so that they could more fully engage in the design process. We designed a potentially reusable solution for achieving SA in multi-stakeholder, multi-UAV, emergency response applications.

cs.HC↗

arXiv2019

Scenarios and Recommendations for Ethical Interpretive AI

Artificially intelligent systems, given a set of non-trivial ethical rules to follow, will inevitably be faced with scenarios which call into question the scope of those rules. In such cases, human reasoners typically will engage in interpretive reasoning, where interpretive arguments are used to support or attack claims that some rule should be understood a certain way. Artificially intelligent reasoners, however, currently lack the ability to carry out human-like interpretive reasoning, and we argue that bridging this gulf is of tremendous importance to human-centered AI. In order to better understand how future artificial reasoners capable of human-like interpretive reasoning must be developed, we have collected a dataset of ethical rules, scenarios designed to invoke interpretive reasoning, and interpretations of those scenarios. We perform a qualitative analysis of our dataset, and summarize our findings in the form of practical recommendations.

cs.AI↗