Related papers: Honeypot Protocol
Black-box safety evaluation of AI systems assumes model behavior on test distributions reliably predicts deployment performance. We formalize and challenge this assumption through latent context-conditioned policies -- models whose outputs…
Autonomous AI agents are being deployed with filesystem access, email control, and multi-step planning. This thesis contributes to four open problems in AI safety: understanding dangerous internal computations, removing dangerous behaviors…
AI pentesting agents are increasingly credible as offensive security systems, but current benchmarks still provide limited guidance on which will perform best in real-world targets. Existing evaluation protocols assess and optimize for…
The proliferation of insecure Internet-connected devices gave rise to the IoT botnets which can grow very large rapidly and may perform high-impact cyber-attacks. The related studies for tackling IoT botnets are concerned with either…
This paper studies binary hypothesis testing based on measurements from a set of sensors, a subset of which can be compromised by an attacker. The measurements from a compromised sensor can be manipulated arbitrarily by the adversary. The…
The model of a non-Bayesian agent who faces a repeated game with incomplete information against Nature is an appropriate tool for modeling general agent-environment interactions. In such a model the environment state (controlled by Nature)…
Learning effective contextual-bandit policies from past actions of a deployed system is highly desirable in many settings (e.g. voice assistants, recommendation, search), since it enables the reuse of large amounts of log data.…
We present a novel class of proof-of-position algorithms: Tree-Proof-of-Position (T-PoP). This algorithm is decentralised, collaborative and can be computed in a privacy preserving manner, such that agents do not need to reveal their…
We consider the contextual bandit problem on general action and context spaces, where the learner's rewards depend on their selected actions and an observable context. This generalizes the standard multi-armed bandit to the case where side…
Industrial Operational Technology (OT) systems are increasingly targeted by cyber-attacks due to their integration with Information Technology (IT) systems in the Industry 4.0 era. Besides intrusion detection systems, honeypots can…
In multi-hop mobile ad hoc networks (MANETs),mobile nodes cooperate with each other without using any infrastructure such as access points or base stations. Security remains a major challenge for these networks due to their features of open…
The steadily increasing level of automation in human-centred systems demands rigorous design methods for analysing and controlling interactions between humans and automated components, especially in safety-critical applications. The…
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…
We introduce the cram method as a general statistical framework for evaluating the final learned policy from a multi-armed contextual bandit algorithm, using the dataset generated by the same bandit algorithm. The proposed on-policy…
In this paper, hypernetworks are trained to generate behaviors across a range of unseen task conditions, via a novel TD-based training objective and data from a set of near-optimal RL solutions for training tasks. This work relates to meta…
Real-world artificial intelligence (AI) systems are increasingly required to operate autonomously in dynamic, uncertain, and continuously changing environments. However, most existing AI models rely on predefined objectives, static training…
Computer-use agents (CUAs) can now autonomously complete complex tasks in real digital environments, but when misled, they can also be used to automate harmful actions programmatically. Existing safety evaluations largely target explicit…
We introduce a multi-turn benchmark for evaluating personalised alignment in LLM-based AI assistants, focusing on their ability to handle user-provided safety-critical contexts. Our assessment of ten leading models across five scenarios…
Operators performing high-stakes, safety-critical tasks - such as air traffic controllers, surgeons, or mission control personnel - must maintain exceptional cognitive performance under variable and often stressful conditions. This paper…
AI systems that learn through reward feedback about the actions they take are increasingly deployed in domains that have significant impact on our daily life. However, in many cases the online rewards should not be the only guiding…