Related papers: Reconciling progress-insensitive noninterference a…

We investigate the combination between causal/zero-delay source coding and information-theoretic secrecy. Two source coding models with secrecy constraints are considered. We start by considering zero-delay perfectly secret lossless…

Departing from traditional communication theory where decoding algorithms are assumed to perform without error, a system where noise perturbs both computational devices and communication channels is considered here. This paper studies…

In quantitative information flow we say that program $Q$ is "at least as secure as" $P$ just when the amount of secret information flowing from $Q$ is never more than flows from $P$, with of course a suitable quantification of "flow". This…

Diffusion language models (DLMs) generate tokens in parallel through iterative denoising, which can reduce latency and enable bidirectional conditioning. However, the safety risks posed by jailbreak attacks that exploit this inference…

In-context learning (ICL) has become one of the most popular learning paradigms. While there is a growing body of literature focusing on prompt engineering, there is a lack of systematic analysis comparing the effects of prompts across…

We propose selective debiasing -- an inference-time safety mechanism designed to enhance the overall model quality in terms of prediction performance and fairness, especially in scenarios where retraining the model is impractical. The…

The rise of "jailbreak" attacks on language models has led to a flurry of defenses aimed at preventing undesirable responses. We critically examine the two stages of the defense pipeline: (i) defining what constitutes unsafe outputs, and…

Manipulations of return addresses on the stack are the basis for a variety of attacks on programs written in memory unsafe languages. Dual stack schemes for protecting return addresses promise an efficient and effective defense against such…

Employees work in increasingly digital environments that enable advanced analytics. Yet, they lack oversight over the systems that process their data. That means that potential analysis errors or hidden biases are hard to uncover. Recent…

A deep learning system typically suffers from a lack of reproducibility that is partially rooted in hardware or software implementation details. The irreproducibility leads to skepticism in deep learning technologies and it can hinder them…

Recent advances in large language models (LLMs) and vision-language models (VLMs) have enabled powerful autonomous agents capable of complex reasoning and multi-modal tool use. Despite their growing capabilities, today's agent frameworks…

With the advent of quantum cloud computing, the security of delegated quantum computation has become of utmost importance. While multiple statistically secure blind verification schemes in the prepare-and-send model have been proposed, none…

Ensuring that information flowing through a network is secure from manipulation and eavesdropping by unauthorized parties is an important task for network administrators. Many cyber attacks rely on a lack of network-level information flow…

Guaranteeing safety in the presence of unmatched disturbances -- uncertainties that cannot be directly canceled by the control input -- remains a key challenge in nonlinear control. This paper presents a constructive approach to…

Denning's lattice model provided secure information flow analyses with an intuitive mathematical foundation: the lattice ordering determines permitted flows. We examine how this framework may be extended to support the flow of information…

While current machine learning models have impressive performance over a wide range of applications, their large size and complexity render them unsuitable for tasks such as remote monitoring on edge devices with limited storage and…

Websites today routinely combine JavaScript from multiple sources, both trusted and untrusted. Hence, JavaScript security is of paramount importance. A specific interesting problem is information flow control (IFC) for JavaScript. In this…

Information fusion deals with the integration and merging of data and information from multiple (heterogeneous) sources. In many cases, the information that needs to be fused has security classification. The result of the fusion process is…

Much research has studied foundations for correct and reliable communication-centric systems. A salient approach to correctness uses session types to enforce structured communications; a recent approach to reliability uses reversible…

Diffusion models struggle to produce samples that respect constraints, a common requirement in scientific applications. Recent approaches have introduced regularization terms in the loss or guidance methods during sampling to enforce such…