Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
arXiv:2603.13275v4 Announce Type: replace-cross Abstract: Accurate prediction of surgical duration is pivotal for hospital resource management. Although recent supervised learning approaches, from ma...
arXiv:2603.18008v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly used for mental-health support; yet prevailing evaluation methods--fluency metrics, preference...
arXiv:2605.06187v2 Announce Type: replace-cross Abstract: Black-box optimization in science and engineering often comes with side information: experts, simulators, pretrained predictors, or heuristic...
arXiv:2605.18991v3 Announce Type: replace-cross Abstract: We take the position that agent security must be approached as a systems problem: the AI model powering the agent must be treated as an untru...
arXiv:2605.21272v2 Announce Type: replace-cross Abstract: Training large text-to-image models requires high-quality, curated datasets with diverse content and detailed captions. Yet the cost and comp...
arXiv:2609.34458v2 Announce Type: replace-cross Abstract: Classical statistical theory is insufficient to explain the successes of general-purpose AI models, because it depends on handcrafted inducti...
arXiv:2611.01680v2 Announce Type: replace-cross Abstract: Reconstructing three-dimensional (3D) complex ocean acoustic fields from sparse measurements is challenging due to highly oscillatory wave st...
arXiv:2611.05528v1 Announce Type: new Abstract: Across a wide range of applications, kernel measures are used to tune hyperparameters, select among candidate kernels, and assess the suitability of ke...
arXiv:2611.05594v1 Announce Type: new Abstract: Relevance propagation, which redistributes a network's prediction backward through the layers, is usually treated as a heuristic. We show that it is th...
arXiv:2611.06082v1 Announce Type: new Abstract: Fast adversarial training (FAT) makes robust training affordable by replacing the multi-step attack with a single-step one, but it suffers from catastr...
arXiv:2611.06194v1 Announce Type: new Abstract: Reliability management of industrial fleets depends on early fault warnings, yet lifecycle data across similar assets are highly imbalanced. A few asse...
arXiv:2611.05580v1 Announce Type: cross Abstract: We study repeated general games over $T$ rounds between a utility-maximizing optimizer and a learning agent, where budget constraints couple the play...
arXiv:2611.06432v1 Announce Type: cross Abstract: Reliable channel estimation and symbol detection in 5G New Radio (NR), as well as emerging 6G radio systems, rely on demodulation reference signals (...
arXiv:2611.06582v1 Announce Type: cross Abstract: Deep neural networks train best when they are initialised at the edge of chaos, the critical line in the plane of weight and bias variances on which...
arXiv:2611.06653v1 Announce Type: cross Abstract: A robot action can satisfy shape and range constraints yet carry the wrong semantics: axes may be permuted, signs inverted, rotations rescaled, or gr...
arXiv:2611.06802v1 Announce Type: cross Abstract: Hamilton-Jacobi (HJ) Reachability identifies the unsafe states of a dynamical system from which it can be forced into failure under worst-case distur...
arXiv:2611.06819v1 Announce Type: cross Abstract: Choosing a quantum error-correcting code for a real device means scoring candidates with circuit-level simulations so slow that only a few hundred ca...
arXiv:2611.06830v1 Announce Type: cross Abstract: Current generative AI alignment relies on a myopic paradigm that prioritizes immediate conversational satisfaction. By relying on short-horizon feedb...
arXiv:2404.17144v3 Announce Type: replace Abstract: The response time of a biosensor is a crucial metric in safety-critical applications such as medical diagnostics where an earlier diagnosis can mar...
arXiv:2502.06774v5 Announce Type: replace Abstract: Ensuring neural networks adhere to domain-specific constraints is crucial for addressing safety and trustworthiness while also enhancing inference...