Related papers: Combining Cost-Constrained Runtime Monitors for AI…
We study AI alignment through the lens of law-and-economics models of deterrence and enforcement. In these models, misconduct is not treated as an external failure, but as a strategic response to incentives: an actor weighs the gain from…
The behavior of neural networks (NNs) on previously unseen types of data (out-of-distribution or OOD) is typically unpredictable. This can be dangerous if the network's output is used for decision-making in a safety-critical system. Hence,…
Control evaluations measure whether monitoring and security protocols for AI systems prevent intentionally subversive AI models from causing harm. Our work presents the first control evaluation performed in an agent environment. We…
We study the problem of policy repair for learning-based control policies in safety-critical settings. We consider an architecture where a high-performance learning-based control policy (e.g. one trained as a neural network) is paired with…
For deep neural networks (DNNs) to be used in safety-critical autonomous driving tasks, it is desirable to monitor in operation time if the input for the DNN is similar to the data used in DNN training. While recent results in monitoring…
Pair programming is a widely used collaborative learning practice in computer science education yet its effectiveness varies substantially due to breakdowns in coordination attention and cognitive regulation between partners. This paper…
The field of AI Control seeks to develop robust control protocols, deployment safeguards for untrusted AI which may be intentionally subversive. However, existing protocols that rely on weaker monitors to detect unsafe behavior often fail…
With the advent of Industry 4.0, industrial facilities and critical infrastructures are transforming into an ecosystem of heterogeneous physical and cyber components, such as programmable logic controllers, increasingly interconnected and…
In cognitive radio networks (CRNs), dynamic spectrum access has been proposed to improve the spectrum utilization, but it also generates spectrum misuse problems. One common solution to these problems is to deploy monitors to detect…
AI coding scaffolds like Claude Code and Codex use retrying: blocking actions flagged as risky and continuing the trajectory. We study retrying from an AI control perspective, which treats the model as potentially adversarial. We find that…
This paper addresses security challenges in multi-robot systems (MRS) where adversaries may compromise robot control, risking unauthorized access to forbidden areas. We propose a novel multi-robot optimal planning algorithm that integrates…
Researchers working on the automatic parallelization of programs have long known that too much parallelism can be even worse for performance than too little, because spawning a task to be run on another CPU incurs overheads.…
Conventional approaches for ensuring the security of application software at run-time, through monitoring, either produce (high rates of) false alarms (e.g. intrusion detection systems) or limit application performance (e.g. run-time…
Physical human-robot interaction (pHRI) integrates the benefits of human operator and a collaborative robot in tasks involving physical interaction, with the aim of increasing the task performance. However, the design of interaction…
The goal of diagnosis is to compute good repair strategies in response to anomalous system behavior. In a decision theoretic framework, a good repair strategy has low expected cost. In a general formulation of the problem, the computation…
Monitoring is an important body of techniques in runtime verification of real-time, embedded, and cyber-physical systems. Mathematically, the monitoring problem can be formalized as a pattern matching problem against a pattern automaton.…
We study the impact of synchronous and asynchronous monitoring instrumentation on runtime overheads in the context of a runtime verification framework for actor-based systems. We show that, in such a context, asynchronous monitoring incurs…
Machine learning applications frequently come with multiple diverse objectives and constraints that can change over time. Accordingly, trained models can be tuned with sets of hyper-parameters that affect their predictive behavior (e.g.,…
The development of safety-critical systems requires the control of hazards that can potentially cause harm. To this end, safety engineers rely during the development phase on architectural solutions, called safety patterns, such as safety…
We describe a shared control methodology that can, without knowledge of the task, be used to improve a human's control of a dynamic system, be used as a training mechanism, and be used in conjunction with Imitation Learning to generate…